id sid tid token lemma pos bracis-19048 1 1 evaluating evaluate VERB bracis-19048 1 2 clustering cluster VERB bracis-19048 1 3 meta meta NOUN bracis-19048 1 4 - - PUNCT bracis-19048 1 5 features feature NOUN bracis-19048 1 6 for for ADP bracis-19048 1 7 classifier classifier NOUN bracis-19048 1 8 recommendation recommendation NOUN bracis-19048 1 9 | | NOUN bracis-19048 1 10 springer springer NOUN bracis-19048 1 11 nature nature NOUN bracis-19048 1 12 link link PROPN bracis-19048 1 13 ( ( PUNCT bracis-19048 1 14 formerly formerly ADV bracis-19048 1 15 springerlink springerlink NOUN bracis-19048 1 16 ) ) PUNCT bracis-19048 1 17 skip skip VERB bracis-19048 1 18 to to ADP bracis-19048 1 19 main main ADJ bracis-19048 1 20 content content NOUN bracis-19048 1 21 advertisement advertisement NOUN bracis-19048 1 22 log log NOUN bracis-19048 1 23 in in ADP bracis-19048 1 24 menu menu NOUN bracis-19048 1 25 find find VERB bracis-19048 1 26 a a DET bracis-19048 1 27 journal journal NOUN bracis-19048 1 28 publish publish VERB bracis-19048 1 29 with with ADP bracis-19048 1 30 us we PRON bracis-19048 1 31 track track VERB bracis-19048 1 32 your your PRON bracis-19048 1 33 research research NOUN bracis-19048 1 34 search search NOUN bracis-19048 1 35 cart cart NOUN bracis-19048 1 36 home home NOUN bracis-19048 1 37 intelligent intelligent ADJ bracis-19048 1 38 systems system NOUN bracis-19048 1 39 conference conference NOUN bracis-19048 1 40 paper paper NOUN bracis-19048 1 41 evaluating evaluating NOUN bracis-19048 1 42 clustering cluster VERB bracis-19048 1 43 meta meta NOUN bracis-19048 1 44 - - PUNCT bracis-19048 1 45 features feature NOUN bracis-19048 1 46 for for ADP bracis-19048 1 47 classifier classifier NOUN bracis-19048 1 48 recommendation recommendation NOUN bracis-19048 1 49 conference conference NOUN bracis-19048 1 50 paper paper NOUN bracis-19048 1 51 first first ADV bracis-19048 1 52 online online ADV bracis-19048 2 1 : : PUNCT bracis-19048 2 2 28 28 NUM bracis-19048 2 3 november november NOUN bracis-19048 2 4 2021 2021 NUM bracis-19048 3 1 pp pp ADP bracis-19048 3 2 453–467 453–467 NUM bracis-19048 3 3 cite cite VERB bracis-19048 3 4 this this DET bracis-19048 3 5 conference conference NOUN bracis-19048 3 6 paper paper NOUN bracis-19048 3 7 access access NOUN bracis-19048 3 8 provided provide VERB bracis-19048 3 9 by by ADP bracis-19048 3 10 university university PROPN bracis-19048 3 11 of of ADP bracis-19048 3 12 notre notre PROPN bracis-19048 3 13 dame dame PROPN bracis-19048 3 14 hesburgh hesburgh PROPN bracis-19048 3 15 library library PROPN bracis-19048 3 16 download download PROPN bracis-19048 3 17 book book NOUN bracis-19048 3 18 pdf pdf PROPN bracis-19048 3 19 download download NOUN bracis-19048 3 20 book book NOUN bracis-19048 3 21 epub epub PROPN bracis-19048 3 22 intelligent intelligent ADJ bracis-19048 3 23 systems system NOUN bracis-19048 3 24 ( ( PUNCT bracis-19048 3 25 bracis bracis NOUN bracis-19048 3 26 2021 2021 NUM bracis-19048 3 27 ) ) PUNCT bracis-19048 3 28 evaluating evaluate VERB bracis-19048 3 29 clustering cluster VERB bracis-19048 3 30 meta meta NOUN bracis-19048 3 31 - - PUNCT bracis-19048 3 32 features feature NOUN bracis-19048 3 33 for for ADP bracis-19048 3 34 classifier classifier NOUN bracis-19048 3 35 recommendation recommendation NOUN bracis-19048 3 36 download download NOUN bracis-19048 3 37 book book NOUN bracis-19048 3 38 pdf pdf PROPN bracis-19048 3 39 download download NOUN bracis-19048 3 40 book book NOUN bracis-19048 3 41 epub epub PROPN bracis-19048 4 1 luís luís PROPN bracis-19048 4 2 p. p. PROPN bracis-19048 4 3 f. f. PROPN bracis-19048 4 4 garcia garcia PROPN bracis-19048 4 5     SPACE bracis-19048 4 6 orcid orcid PROPN bracis-19048 4 7 : : PUNCT bracis-19048 4 8 orcid.org/0000-0003-0679-914310 orcid.org/0000-0003-0679-914310 ADJ bracis-19048 4 9 , , PUNCT bracis-19048 4 10 felipe felipe PROPN bracis-19048 4 11 campelo campelo PROPN bracis-19048 4 12     SPACE bracis-19048 4 13 orcid orcid NOUN bracis-19048 4 14 : : PUNCT bracis-19048 4 15 orcid.org/0000-0001-8432-432511 orcid.org/0000-0001-8432-432511 NOUN bracis-19048 4 16 , , PUNCT bracis-19048 4 17 guilherme guilherme PROPN bracis-19048 4 18 n. n. PROPN bracis-19048 4 19 ramos ramos PROPN bracis-19048 4 20     SPACE bracis-19048 4 21 orcid orcid NOUN bracis-19048 4 22 : : PUNCT bracis-19048 4 23 orcid.org/0000-0001-5859-736210 orcid.org/0000-0001-5859-736210 ADJ bracis-19048 4 24 , , PUNCT bracis-19048 4 25 adriano adriano PROPN bracis-19048 4 26 rivolli rivolli PROPN bracis-19048 4 27     SPACE bracis-19048 4 28 orcid orcid PROPN bracis-19048 4 29 : : PUNCT bracis-19048 4 30 orcid.org/0000-0001-6445-300712 orcid.org/0000-0001-6445-300712 NOUN bracis-19048 4 31 & & CCONJ bracis-19048 4 32 … … PUNCT bracis-19048 4 33 andré andré PROPN bracis-19048 4 34 c. c. PROPN bracis-19048 4 35 p. p. PROPN bracis-19048 4 36 de de PROPN bracis-19048 4 37 l. l. PROPN bracis-19048 4 38 f. f. PROPN bracis-19048 4 39 de de PROPN bracis-19048 4 40 carvalho carvalho PROPN bracis-19048 4 41     SPACE bracis-19048 4 42 orcid orcid NOUN bracis-19048 4 43 : : PUNCT bracis-19048 4 44 orcid.org/0000-0002-4765-645913 orcid.org/0000-0002-4765-645913 PROPN bracis-19048 4 45     SPACE bracis-19048 4 46 show show VERB bracis-19048 4 47 authors author NOUN bracis-19048 4 48 part part NOUN bracis-19048 4 49 of of ADP bracis-19048 4 50 the the DET bracis-19048 4 51 book book NOUN bracis-19048 4 52 series series NOUN bracis-19048 4 53 : : PUNCT bracis-19048 4 54 lecture lecture NOUN bracis-19048 4 55 notes note NOUN bracis-19048 4 56 in in ADP bracis-19048 4 57 computer computer NOUN bracis-19048 4 58 science science NOUN bracis-19048 4 59 ( ( PUNCT bracis-19048 4 60 ( ( PUNCT bracis-19048 4 61 lnai lnai ADJ bracis-19048 4 62 , , PUNCT bracis-19048 4 63 volume volume NOUN bracis-19048 4 64 13073 13073 NUM bracis-19048 4 65 ) ) PUNCT bracis-19048 4 66 ) ) PUNCT bracis-19048 4 67 included include VERB bracis-19048 4 68 in in ADP bracis-19048 4 69 the the DET bracis-19048 4 70 following follow VERB bracis-19048 4 71 conference conference NOUN bracis-19048 4 72 series series NOUN bracis-19048 4 73 : : PUNCT bracis-19048 4 74 brazilian brazilian ADJ bracis-19048 4 75 conference conference NOUN bracis-19048 4 76 on on ADP bracis-19048 4 77 intelligent intelligent ADJ bracis-19048 4 78 systems system NOUN bracis-19048 4 79 802 802 NUM bracis-19048 4 80 accesses access NOUN bracis-19048 4 81 1 1 NUM bracis-19048 4 82 citation citation NOUN bracis-19048 4 83 1 1 NUM bracis-19048 4 84 altmetric altmetric ADJ bracis-19048 4 85 abstract abstract ADJ bracis-19048 4 86 data datum NOUN bracis-19048 4 87 availability availability NOUN bracis-19048 4 88 in in ADP bracis-19048 4 89 a a DET bracis-19048 4 90 wide wide ADJ bracis-19048 4 91 variety variety NOUN bracis-19048 4 92 of of ADP bracis-19048 4 93 domains domain NOUN bracis-19048 4 94 has have AUX bracis-19048 4 95 boosted boost VERB bracis-19048 4 96 the the DET bracis-19048 4 97 use use NOUN bracis-19048 4 98 of of ADP bracis-19048 4 99 machine machine NOUN bracis-19048 4 100 learning learn VERB bracis-19048 4 101 techniques technique NOUN bracis-19048 4 102 for for ADP bracis-19048 4 103 knowledge knowledge NOUN bracis-19048 4 104 discovery discovery NOUN bracis-19048 4 105 and and CCONJ bracis-19048 4 106 classification classification NOUN bracis-19048 4 107 . . PUNCT bracis-19048 5 1 the the DET bracis-19048 5 2 performance performance NOUN bracis-19048 5 3 of of ADP bracis-19048 5 4 a a DET bracis-19048 5 5 technique technique NOUN bracis-19048 5 6 in in ADP bracis-19048 5 7 a a DET bracis-19048 5 8 given give VERB bracis-19048 5 9 classification classification NOUN bracis-19048 5 10 task task NOUN bracis-19048 5 11 is be AUX bracis-19048 5 12 significantly significantly ADV bracis-19048 5 13 impacted impact VERB bracis-19048 5 14 by by ADP bracis-19048 5 15 specific specific ADJ bracis-19048 5 16 characteristics characteristic NOUN bracis-19048 5 17 of of ADP bracis-19048 5 18 the the DET bracis-19048 5 19 dataset dataset NOUN bracis-19048 5 20 , , PUNCT bracis-19048 5 21 which which PRON bracis-19048 5 22 makes make VERB bracis-19048 5 23 the the DET bracis-19048 5 24 problem problem NOUN bracis-19048 5 25 of of ADP bracis-19048 5 26 choosing choose VERB bracis-19048 5 27 the the DET bracis-19048 5 28 most most ADV bracis-19048 5 29 adequate adequate ADJ bracis-19048 5 30 approach approach NOUN bracis-19048 5 31 a a DET bracis-19048 5 32 challenging challenging ADJ bracis-19048 5 33 one one NUM bracis-19048 5 34 . . PUNCT bracis-19048 6 1 meta meta ADJ bracis-19048 6 2 - - PUNCT bracis-19048 6 3 learning learn VERB bracis-19048 6 4 approaches approach NOUN bracis-19048 6 5 , , PUNCT bracis-19048 6 6 which which PRON bracis-19048 6 7 learn learn VERB bracis-19048 6 8 from from ADP bracis-19048 6 9 meta meta ADJ bracis-19048 6 10 - - PUNCT bracis-19048 6 11 features feature NOUN bracis-19048 6 12 calculated calculate VERB bracis-19048 6 13 from from ADP bracis-19048 6 14 the the DET bracis-19048 6 15 dataset dataset NOUN bracis-19048 6 16 , , PUNCT bracis-19048 6 17 have have AUX bracis-19048 6 18 been be AUX bracis-19048 6 19 successfully successfully ADV bracis-19048 6 20 used use VERB bracis-19048 6 21 to to PART bracis-19048 6 22 suggest suggest VERB bracis-19048 6 23 the the DET bracis-19048 6 24 most most ADV bracis-19048 6 25 suitable suitable ADJ bracis-19048 6 26 classification classification NOUN bracis-19048 6 27 algorithms algorithm NOUN bracis-19048 6 28 for for ADP bracis-19048 6 29 specific specific ADJ bracis-19048 6 30 datasets dataset NOUN bracis-19048 6 31 . . PUNCT bracis-19048 7 1 this this DET bracis-19048 7 2 work work NOUN bracis-19048 7 3 proposes propose VERB bracis-19048 7 4 the the DET bracis-19048 7 5 adaptation adaptation NOUN bracis-19048 7 6 of of ADP bracis-19048 7 7 clustering cluster VERB bracis-19048 7 8 measures measure NOUN bracis-19048 7 9 based base VERB bracis-19048 7 10 on on ADP bracis-19048 7 11 internal internal ADJ bracis-19048 7 12 indices index NOUN bracis-19048 7 13 for for ADP bracis-19048 7 14 supervised supervised ADJ bracis-19048 7 15 problems problem NOUN bracis-19048 7 16 as as ADP bracis-19048 7 17 additional additional ADJ bracis-19048 7 18 meta meta NOUN bracis-19048 7 19 - - PUNCT bracis-19048 7 20 features feature NOUN bracis-19048 7 21 in in ADP bracis-19048 7 22 the the DET bracis-19048 7 23 process process NOUN bracis-19048 7 24 of of ADP bracis-19048 7 25 learning learn VERB bracis-19048 7 26 a a DET bracis-19048 7 27 recommendation recommendation NOUN bracis-19048 7 28 system system NOUN bracis-19048 7 29 for for ADP bracis-19048 7 30 classification classification NOUN bracis-19048 7 31 tasks task NOUN bracis-19048 7 32 . . PUNCT bracis-19048 8 1 the the DET bracis-19048 8 2 gains gain NOUN bracis-19048 8 3 in in ADP bracis-19048 8 4 performance performance NOUN bracis-19048 8 5 due due ADP bracis-19048 8 6 to to ADP bracis-19048 8 7 meta meta VERB bracis-19048 8 8 - - PUNCT bracis-19048 8 9 learning learning NOUN bracis-19048 8 10 and and CCONJ bracis-19048 8 11 the the DET bracis-19048 8 12 additional additional ADJ bracis-19048 8 13 meta meta NOUN bracis-19048 8 14 - - PUNCT bracis-19048 8 15 features feature NOUN bracis-19048 8 16 are be AUX bracis-19048 8 17 investigated investigate VERB bracis-19048 8 18 with with ADP bracis-19048 8 19 experiments experiment NOUN bracis-19048 8 20 based base VERB bracis-19048 8 21 on on ADP bracis-19048 8 22 400 400 NUM bracis-19048 8 23 datasets dataset NOUN bracis-19048 8 24 , , PUNCT bracis-19048 8 25 representing represent VERB bracis-19048 8 26 diverse diverse ADJ bracis-19048 8 27 application application NOUN bracis-19048 8 28 contexts context NOUN bracis-19048 8 29 and and CCONJ bracis-19048 8 30 domains domain NOUN bracis-19048 8 31 . . PUNCT bracis-19048 9 1 results result NOUN bracis-19048 9 2 suggest suggest VERB bracis-19048 9 3 that that SCONJ bracis-19048 9 4 ( ( PUNCT bracis-19048 9 5 i i NOUN bracis-19048 9 6 ) ) PUNCT bracis-19048 9 7 meta meta VERB bracis-19048 9 8 - - PUNCT bracis-19048 9 9 learning learning NOUN bracis-19048 9 10 is be AUX bracis-19048 9 11 a a DET bracis-19048 9 12 viable viable ADJ bracis-19048 9 13 solution solution NOUN bracis-19048 9 14 for for ADP bracis-19048 9 15 recommending recommend VERB bracis-19048 9 16 a a DET bracis-19048 9 17 classifier classifier NOUN bracis-19048 9 18 , , PUNCT bracis-19048 9 19 ( ( PUNCT bracis-19048 9 20 ii ii NOUN bracis-19048 9 21 ) ) PUNCT bracis-19048 9 22 the the DET bracis-19048 9 23 use use NOUN bracis-19048 9 24 of of ADP bracis-19048 9 25 clustering clustering ADJ bracis-19048 9 26 features feature NOUN bracis-19048 9 27 can can AUX bracis-19048 9 28 contribute contribute VERB bracis-19048 9 29 to to ADP bracis-19048 9 30 the the DET bracis-19048 9 31 performance performance NOUN bracis-19048 9 32 of of ADP bracis-19048 9 33 the the DET bracis-19048 9 34 recommendation recommendation NOUN bracis-19048 9 35 system system NOUN bracis-19048 9 36 , , PUNCT bracis-19048 9 37 and and CCONJ bracis-19048 9 38 ( ( PUNCT bracis-19048 9 39 iii iii X bracis-19048 9 40 ) ) PUNCT bracis-19048 9 41 the the DET bracis-19048 9 42 computational computational ADJ bracis-19048 9 43 cost cost NOUN bracis-19048 9 44 of of ADP bracis-19048 9 45 meta meta NOUN bracis-19048 9 46 - - PUNCT bracis-19048 9 47 learning learning NOUN bracis-19048 9 48 is be AUX bracis-19048 9 49 substantially substantially ADV bracis-19048 9 50 smaller small ADJ bracis-19048 9 51 than than ADP bracis-19048 9 52 that that PRON bracis-19048 9 53 of of ADP bracis-19048 9 54 running run VERB bracis-19048 9 55 all all DET bracis-19048 9 56 candidate candidate NOUN bracis-19048 9 57 classifiers classifier NOUN bracis-19048 9 58 in in ADP bracis-19048 9 59 order order NOUN bracis-19048 9 60 to to PART bracis-19048 9 61 select select VERB bracis-19048 9 62 the the DET bracis-19048 9 63 best good ADJ bracis-19048 9 64 . . PUNCT bracis-19048 10 1 access access NOUN bracis-19048 10 2 provided provide VERB bracis-19048 10 3 by by ADP bracis-19048 10 4 university university PROPN bracis-19048 10 5 of of ADP bracis-19048 10 6 notre notre PROPN bracis-19048 10 7 dame dame PROPN bracis-19048 10 8 hesburgh hesburgh PROPN bracis-19048 10 9 library library PROPN bracis-19048 10 10 . . PUNCT bracis-19048 11 1 download download PROPN bracis-19048 11 2 conference conference NOUN bracis-19048 11 3 paper paper NOUN bracis-19048 11 4 pdf pdf NOUN bracis-19048 11 5 similar similar ADJ bracis-19048 11 6 content content NOUN bracis-19048 11 7 being be AUX bracis-19048 11 8 viewed view VERB bracis-19048 11 9 by by ADP bracis-19048 11 10 others other NOUN bracis-19048 11 11 evaluating evaluate VERB bracis-19048 11 12 data datum NOUN bracis-19048 11 13 characterization characterization NOUN bracis-19048 11 14 measures measure NOUN bracis-19048 11 15 for for ADP bracis-19048 11 16     SPACE bracis-19048 11 17 clustering cluster VERB bracis-19048 11 18 problems problem NOUN bracis-19048 11 19 in in ADP bracis-19048 11 20     SPACE bracis-19048 11 21 meta meta ADV bracis-19048 11 22 - - PUNCT bracis-19048 11 23 learning learn VERB bracis-19048 11 24 chapter chapter NOUN bracis-19048 11 25 © © NOUN bracis-19048 11 26 2021 2021 NUM bracis-19048 11 27 evolving evolve VERB bracis-19048 11 28 learners learner NOUN bracis-19048 11 29 ’ ' PUNCT bracis-19048 11 30 behavior behavior NOUN bracis-19048 11 31 in in ADP bracis-19048 11 32 data data NOUN bracis-19048 11 33 mining mining NOUN bracis-19048 11 34 article article NOUN bracis-19048 11 35 21 21 NUM bracis-19048 11 36 september september PROPN bracis-19048 11 37 2016 2016 NUM bracis-19048 11 38 dataset dataset NOUN bracis-19048 11 39 characteristics characteristic NOUN bracis-19048 11 40 ( ( PUNCT bracis-19048 11 41 metafeatures metafeature NOUN bracis-19048 11 42 ) ) PUNCT bracis-19048 11 43 chapter chapter NOUN bracis-19048 11 44 © © PROPN bracis-19048 11 45 2022 2022 NUM bracis-19048 11 46 explore explore VERB bracis-19048 11 47 related relate VERB bracis-19048 11 48 subjects subject NOUN bracis-19048 11 49 discover discover VERB bracis-19048 11 50 the the DET bracis-19048 11 51 latest late ADJ bracis-19048 11 52 articles article NOUN bracis-19048 11 53 , , PUNCT bracis-19048 11 54 books book NOUN bracis-19048 11 55 and and CCONJ bracis-19048 11 56 news news NOUN bracis-19048 11 57 in in ADP bracis-19048 11 58 related related ADJ bracis-19048 11 59 subjects subject NOUN bracis-19048 11 60 , , PUNCT bracis-19048 11 61 suggested suggest VERB bracis-19048 11 62 using use VERB bracis-19048 11 63 machine machine NOUN bracis-19048 11 64 learning learning NOUN bracis-19048 11 65 . . PUNCT bracis-19048 12 1 categorization categorization NOUN bracis-19048 12 2 data datum NOUN bracis-19048 12 3 mining mining NOUN bracis-19048 12 4 and and CCONJ bracis-19048 12 5 knowledge knowledge NOUN bracis-19048 12 6 discovery discovery PROPN bracis-19048 12 7 data data PROPN bracis-19048 12 8 mining mining NOUN bracis-19048 12 9 learning learn VERB bracis-19048 12 10 algorithms algorithm NOUN bracis-19048 12 11 machine machine NOUN bracis-19048 12 12 learning learn VERB bracis-19048 12 13 statistical statistical ADJ bracis-19048 12 14 learning learn VERB bracis-19048 12 15 1 1 NUM bracis-19048 12 16 introduction introduction NOUN bracis-19048 12 17 data datum NOUN bracis-19048 12 18 analysis analysis NOUN bracis-19048 12 19 is be AUX bracis-19048 12 20 a a DET bracis-19048 12 21 pervasive pervasive ADJ bracis-19048 12 22 activity activity NOUN bracis-19048 12 23 in in ADP bracis-19048 12 24 several several ADJ bracis-19048 12 25 areas area NOUN bracis-19048 12 26 of of ADP bracis-19048 12 27 industrial industrial ADJ bracis-19048 12 28 and and CCONJ bracis-19048 12 29 scientific scientific ADJ bracis-19048 12 30 activities activity NOUN bracis-19048 12 31 , , PUNCT bracis-19048 12 32 often often ADV bracis-19048 12 33 involving involve VERB bracis-19048 12 34 machine machine NOUN bracis-19048 12 35 learning learning NOUN bracis-19048 12 36 ( ( PUNCT bracis-19048 12 37 ml ml NOUN bracis-19048 12 38 ) ) PUNCT bracis-19048 12 39 approaches approach NOUN bracis-19048 12 40 , , PUNCT bracis-19048 12 41 which which PRON bracis-19048 12 42 have have AUX bracis-19048 12 43 been be AUX bracis-19048 12 44 successfully successfully ADV bracis-19048 12 45 applied apply VERB bracis-19048 12 46 in in ADP bracis-19048 12 47 a a DET bracis-19048 12 48 wide wide ADJ bracis-19048 12 49 variety variety NOUN bracis-19048 12 50 of of ADP bracis-19048 12 51 domains domain NOUN bracis-19048 12 52     SPACE bracis-19048 12 53 [ [ X bracis-19048 12 54 35 35 NUM bracis-19048 12 55 ] ] PUNCT bracis-19048 12 56 . . PUNCT bracis-19048 13 1 however however ADV bracis-19048 13 2 , , PUNCT bracis-19048 13 3 the the DET bracis-19048 13 4 variability variability NOUN bracis-19048 13 5 of of ADP bracis-19048 13 6 dataset dataset ADJ bracis-19048 13 7 characteristics characteristic NOUN bracis-19048 13 8 makes make VERB bracis-19048 13 9 it it PRON bracis-19048 13 10 challenging challenge VERB bracis-19048 13 11 to to PART bracis-19048 13 12 accurately accurately ADV bracis-19048 13 13 select select VERB bracis-19048 13 14 the the DET bracis-19048 13 15 most most ADV bracis-19048 13 16 adequate adequate ADJ bracis-19048 13 17 ml ml PART bracis-19048 13 18 approach approach NOUN bracis-19048 13 19 to to PART bracis-19048 13 20 handle handle VERB bracis-19048 13 21 new new ADJ bracis-19048 13 22 data datum NOUN bracis-19048 13 23 . . PUNCT bracis-19048 14 1 this this DET bracis-19048 14 2 issue issue NOUN bracis-19048 14 3 has have AUX bracis-19048 14 4 also also ADV bracis-19048 14 5 been be AUX bracis-19048 14 6 tackled tackle VERB bracis-19048 14 7 from from ADP bracis-19048 14 8 a a DET bracis-19048 14 9 ml ml NOUN bracis-19048 14 10 perspective perspective NOUN bracis-19048 14 11 , , PUNCT bracis-19048 14 12 where where SCONJ bracis-19048 14 13 the the DET bracis-19048 14 14 task task NOUN bracis-19048 14 15 is be AUX bracis-19048 14 16 learning learn VERB bracis-19048 14 17 , , PUNCT bracis-19048 14 18 based base VERB bracis-19048 14 19 on on ADP bracis-19048 14 20 meta meta ADJ bracis-19048 14 21 - - PUNCT bracis-19048 14 22 data datum NOUN bracis-19048 14 23 extracted extract VERB bracis-19048 14 24 from from ADP bracis-19048 14 25 several several ADJ bracis-19048 14 26 datasets dataset NOUN bracis-19048 14 27 related relate VERB bracis-19048 14 28 to to ADP bracis-19048 14 29 a a DET bracis-19048 14 30 given give VERB bracis-19048 14 31 application application NOUN bracis-19048 14 32 or or CCONJ bracis-19048 14 33 domain domain NOUN bracis-19048 14 34 , , PUNCT bracis-19048 14 35 a a DET bracis-19048 14 36 model model NOUN bracis-19048 14 37 to to PART bracis-19048 14 38 predict predict VERB bracis-19048 14 39 which which DET bracis-19048 14 40 ml ml NOUN bracis-19048 14 41 approach approach NOUN bracis-19048 14 42 may may AUX bracis-19048 14 43 be be AUX bracis-19048 14 44 better well ADV bracis-19048 14 45 suited suit VERB bracis-19048 14 46 for for ADP bracis-19048 14 47 exploring explore VERB bracis-19048 14 48 that that DET bracis-19048 14 49 particular particular ADJ bracis-19048 14 50 domain domain NOUN bracis-19048 14 51 . . PUNCT bracis-19048 15 1 this this DET bracis-19048 15 2 approach approach NOUN bracis-19048 15 3 , , PUNCT bracis-19048 15 4 commonly commonly ADV bracis-19048 15 5 referred refer VERB bracis-19048 15 6 to to ADP bracis-19048 15 7 as as ADP bracis-19048 15 8 meta meta ADV bracis-19048 15 9 - - PUNCT bracis-19048 15 10 learning learn VERB bracis-19048 15 11 ( ( PUNCT bracis-19048 15 12 mtl mtl PROPN bracis-19048 15 13 ) ) PUNCT bracis-19048 15 14 , , PUNCT bracis-19048 15 15 has have AUX bracis-19048 15 16 shown show VERB bracis-19048 15 17 promising promising ADJ bracis-19048 15 18 results result NOUN bracis-19048 15 19 in in ADP bracis-19048 15 20 predicting predict VERB bracis-19048 15 21 classifier classifier NOUN bracis-19048 15 22 performances performance NOUN bracis-19048 15 23 for for ADP bracis-19048 15 24 a a DET bracis-19048 15 25 problem problem NOUN bracis-19048 15 26 , , PUNCT bracis-19048 15 27 based base VERB bracis-19048 15 28 on on ADP bracis-19048 15 29 statistical statistical ADJ bracis-19048 15 30 or or CCONJ bracis-19048 15 31 geometrical geometrical ADJ bracis-19048 15 32 characteristics characteristic NOUN bracis-19048 15 33 of of ADP bracis-19048 15 34 a a DET bracis-19048 15 35 dataset dataset NOUN bracis-19048 15 36     SPACE bracis-19048 16 1 [ [ X bracis-19048 16 2 16 16 NUM bracis-19048 16 3 ] ] PUNCT bracis-19048 16 4 . . PUNCT bracis-19048 17 1 mtl mtl PROPN bracis-19048 17 2 approaches approach NOUN bracis-19048 17 3 consist consist VERB bracis-19048 17 4 of of ADP bracis-19048 17 5 the the DET bracis-19048 17 6 data data NOUN bracis-19048 17 7 - - PUNCT bracis-19048 17 8 driven drive VERB bracis-19048 17 9 selection selection NOUN bracis-19048 17 10 of of ADP bracis-19048 17 11 techniques technique NOUN bracis-19048 17 12 , , PUNCT bracis-19048 17 13 through through ADP bracis-19048 17 14 knowledge knowledge NOUN bracis-19048 17 15 extracted extract VERB bracis-19048 17 16 from from ADP bracis-19048 17 17 previous previous ADJ bracis-19048 17 18 tasks task NOUN bracis-19048 17 19 , , PUNCT bracis-19048 17 20 which which PRON bracis-19048 17 21 may may AUX bracis-19048 17 22 be be AUX bracis-19048 17 23 then then ADV bracis-19048 17 24 applied apply VERB bracis-19048 17 25 by by ADP bracis-19048 17 26 a a DET bracis-19048 17 27 recommendation recommendation NOUN bracis-19048 17 28 system system NOUN bracis-19048 17 29 to to PART bracis-19048 17 30 predict predict VERB bracis-19048 17 31 the the DET bracis-19048 17 32 preferred preferred ADJ bracis-19048 17 33 approach approach NOUN bracis-19048 17 34 for for ADP bracis-19048 17 35 a a DET bracis-19048 17 36 new new ADJ bracis-19048 17 37 , , PUNCT bracis-19048 17 38 previously previously ADV bracis-19048 17 39 unseen unseen ADJ bracis-19048 17 40 problem problem NOUN bracis-19048 17 41     SPACE bracis-19048 18 1 [ [ X bracis-19048 18 2 4 4 NUM bracis-19048 18 3 ] ] PUNCT bracis-19048 18 4 . . PUNCT bracis-19048 19 1 this this PRON bracis-19048 19 2 depends depend VERB bracis-19048 19 3 on on ADP bracis-19048 19 4 the the DET bracis-19048 19 5 construction construction NOUN bracis-19048 19 6 of of ADP bracis-19048 19 7 a a DET bracis-19048 19 8 meta meta ADV bracis-19048 19 9 - - PUNCT bracis-19048 19 10 dataset dataset NOUN bracis-19048 19 11 from from ADP bracis-19048 19 12 the the DET bracis-19048 19 13 information information NOUN bracis-19048 19 14 on on ADP bracis-19048 19 15 a a DET bracis-19048 19 16 group group NOUN bracis-19048 19 17 of of ADP bracis-19048 19 18 datasets dataset NOUN bracis-19048 19 19 in in ADP bracis-19048 19 20 the the DET bracis-19048 19 21 class class NOUN bracis-19048 19 22 of of ADP bracis-19048 19 23 the the DET bracis-19048 19 24 problem problem NOUN bracis-19048 19 25 . . PUNCT bracis-19048 20 1 for for ADP bracis-19048 20 2 each each DET bracis-19048 20 3 dataset dataset NOUN bracis-19048 20 4 in in ADP bracis-19048 20 5 the the DET bracis-19048 20 6 group group NOUN bracis-19048 20 7 , , PUNCT bracis-19048 20 8 its its PRON bracis-19048 20 9 descriptive descriptive ADJ bracis-19048 20 10 characteristics characteristic NOUN bracis-19048 20 11 ( ( PUNCT bracis-19048 20 12 meta meta X bracis-19048 20 13 - - PUNCT bracis-19048 20 14 features feature NOUN bracis-19048 20 15 ) ) PUNCT bracis-19048 20 16 are be AUX bracis-19048 20 17 extracted extract VERB bracis-19048 20 18 and and CCONJ bracis-19048 20 19 combined combine VERB bracis-19048 20 20 into into ADP bracis-19048 20 21 one one NUM bracis-19048 20 22 or or CCONJ bracis-19048 20 23 more more ADJ bracis-19048 20 24 meta meta ADJ bracis-19048 20 25 - - PUNCT bracis-19048 20 26 examples example NOUN bracis-19048 20 27 , , PUNCT bracis-19048 20 28 which which PRON bracis-19048 20 29 are be AUX bracis-19048 20 30 then then ADV bracis-19048 20 31 labeled label VERB bracis-19048 20 32 according accord VERB bracis-19048 20 33 to to ADP bracis-19048 20 34 the the DET bracis-19048 20 35 observed observe VERB bracis-19048 20 36 performances performance NOUN bracis-19048 20 37 of of ADP bracis-19048 20 38 different different ADJ bracis-19048 20 39 ml ml NOUN bracis-19048 20 40 algorithms algorithm NOUN bracis-19048 20 41 or or CCONJ bracis-19048 20 42 an an DET bracis-19048 20 43 order order NOUN bracis-19048 20 44 - - PUNCT bracis-19048 20 45 preserving preserve VERB bracis-19048 20 46 transformation transformation NOUN bracis-19048 20 47 such such ADJ bracis-19048 20 48 as as ADP bracis-19048 20 49 their their PRON bracis-19048 20 50 rank rank NOUN bracis-19048 20 51     SPACE bracis-19048 21 1 [ [ X bracis-19048 21 2 4 4 NUM bracis-19048 21 3 ] ] PUNCT bracis-19048 21 4 , , PUNCT bracis-19048 21 5 composing compose VERB bracis-19048 21 6 the the DET bracis-19048 21 7 target target NOUN bracis-19048 21 8 feature feature NOUN bracis-19048 21 9 to to PART bracis-19048 21 10 be be AUX bracis-19048 21 11 predicted predict VERB bracis-19048 21 12 . . PUNCT bracis-19048 22 1 from from ADP bracis-19048 22 2 the the DET bracis-19048 22 3 meta meta ADJ bracis-19048 22 4 - - PUNCT bracis-19048 22 5 dataset dataset NOUN bracis-19048 22 6 , , PUNCT bracis-19048 22 7 a a DET bracis-19048 22 8 meta meta NOUN bracis-19048 22 9 - - PUNCT bracis-19048 22 10 model model NOUN bracis-19048 22 11 can can AUX bracis-19048 22 12 be be AUX bracis-19048 22 13 induced induce VERB bracis-19048 22 14 by by ADP bracis-19048 22 15 a a DET bracis-19048 22 16 learning learn VERB bracis-19048 22 17 algorithm algorithm NOUN bracis-19048 22 18 and and CCONJ bracis-19048 22 19 then then ADV bracis-19048 22 20 used use VERB bracis-19048 22 21 in in ADP bracis-19048 22 22 a a DET bracis-19048 22 23 recommendation recommendation NOUN bracis-19048 22 24 system system NOUN bracis-19048 22 25 to to PART bracis-19048 22 26 predict predict VERB bracis-19048 22 27 which which DET bracis-19048 22 28 algorithm algorithm NOUN bracis-19048 22 29 is be AUX bracis-19048 22 30 expected expect VERB bracis-19048 22 31 to to PART bracis-19048 22 32 have have VERB bracis-19048 22 33 the the DET bracis-19048 22 34 best good ADJ bracis-19048 22 35 performance performance NOUN bracis-19048 22 36 when when SCONJ bracis-19048 22 37 a a DET bracis-19048 22 38 new new ADJ bracis-19048 22 39 problem problem NOUN bracis-19048 22 40 needs need VERB bracis-19048 22 41 to to PART bracis-19048 22 42 be be AUX bracis-19048 22 43 addressed address VERB bracis-19048 22 44     SPACE bracis-19048 23 1 [ [ X bracis-19048 23 2 34 34 NUM bracis-19048 23 3 ] ] PUNCT bracis-19048 23 4 . . PUNCT bracis-19048 24 1 despite despite SCONJ bracis-19048 24 2 their their PRON bracis-19048 24 3 success success NOUN bracis-19048 24 4 , , PUNCT bracis-19048 24 5 most most ADJ bracis-19048 24 6 mtl mtl PROPN bracis-19048 24 7 studies study NOUN bracis-19048 24 8 still still ADV bracis-19048 24 9 lack lack VERB bracis-19048 24 10 an an DET bracis-19048 24 11 in in ADP bracis-19048 24 12 - - PUNCT bracis-19048 24 13 depth depth NOUN bracis-19048 24 14 analysis analysis NOUN bracis-19048 24 15 of of ADP bracis-19048 24 16 the the DET bracis-19048 24 17 meta meta NOUN bracis-19048 24 18 - - PUNCT bracis-19048 24 19 features feature NOUN bracis-19048 24 20     SPACE bracis-19048 25 1 [ [ X bracis-19048 25 2 29 29 NUM bracis-19048 25 3 ] ] PUNCT bracis-19048 25 4 which which PRON bracis-19048 25 5 are be AUX bracis-19048 25 6 known know VERB bracis-19048 25 7 to to PART bracis-19048 25 8 be be AUX bracis-19048 25 9 crucial crucial ADJ bracis-19048 25 10 in in ADP bracis-19048 25 11 the the DET bracis-19048 25 12 successful successful ADJ bracis-19048 25 13 use use NOUN bracis-19048 25 14 of of ADP bracis-19048 25 15 mtl mtl PROPN bracis-19048 25 16     SPACE bracis-19048 26 1 [ [ X bracis-19048 26 2 3 3 NUM bracis-19048 26 3 ] ] PUNCT bracis-19048 26 4 . . PUNCT bracis-19048 27 1 several several ADJ bracis-19048 27 2 works work NOUN bracis-19048 27 3 propose propose VERB bracis-19048 27 4 new new ADJ bracis-19048 27 5 meta meta NOUN bracis-19048 27 6 - - PUNCT bracis-19048 27 7 features feature NOUN bracis-19048 27 8     SPACE bracis-19048 28 1 [ [ X bracis-19048 28 2 15,16,17 15,16,17 NUM bracis-19048 28 3 ] ] PUNCT bracis-19048 28 4 , , PUNCT bracis-19048 28 5 but but CCONJ bracis-19048 28 6 only only ADV bracis-19048 28 7 a a DET bracis-19048 28 8 few few ADJ bracis-19048 28 9 present present ADJ bracis-19048 28 10 important important ADJ bracis-19048 28 11 details detail NOUN bracis-19048 28 12 such such ADJ bracis-19048 28 13 as as ADP bracis-19048 28 14 their their PRON bracis-19048 28 15 asymptotic asymptotic ADJ bracis-19048 28 16 computational computational ADJ bracis-19048 28 17 cost cost NOUN bracis-19048 28 18 , , PUNCT bracis-19048 28 19 the the DET bracis-19048 28 20 degree degree NOUN bracis-19048 28 21 of of ADP bracis-19048 28 22 information information NOUN bracis-19048 28 23 presented present VERB bracis-19048 28 24 , , PUNCT bracis-19048 28 25 or or CCONJ bracis-19048 28 26 the the DET bracis-19048 28 27 importance importance NOUN bracis-19048 28 28 of of ADP bracis-19048 28 29 the the DET bracis-19048 28 30 meta meta NOUN bracis-19048 28 31 - - PUNCT bracis-19048 28 32 features feature NOUN bracis-19048 28 33 for for ADP bracis-19048 28 34 the the DET bracis-19048 28 35 investigated investigate VERB bracis-19048 28 36 problems problem NOUN bracis-19048 28 37     SPACE bracis-19048 29 1 [ [ X bracis-19048 29 2 16 16 NUM bracis-19048 29 3 , , PUNCT bracis-19048 29 4 24 24 NUM bracis-19048 29 5 ] ] PUNCT bracis-19048 29 6 . . PUNCT bracis-19048 30 1 this this DET bracis-19048 30 2 paper paper NOUN bracis-19048 30 3 presents present VERB bracis-19048 30 4 the the DET bracis-19048 30 5 use use NOUN bracis-19048 30 6 of of ADP bracis-19048 30 7 clustering clustering ADJ bracis-19048 30 8 measures measure NOUN bracis-19048 30 9 as as ADP bracis-19048 30 10 meta meta NOUN bracis-19048 30 11 - - PUNCT bracis-19048 30 12 features feature NOUN bracis-19048 30 13 in in ADP bracis-19048 30 14 an an DET bracis-19048 30 15 mtl mtl PROPN bracis-19048 30 16 framework framework NOUN bracis-19048 30 17 , , PUNCT bracis-19048 30 18 to to PART bracis-19048 30 19 learn learn VERB bracis-19048 30 20 a a DET bracis-19048 30 21 recommendation recommendation NOUN bracis-19048 30 22 system system NOUN bracis-19048 30 23 for for ADP bracis-19048 30 24 classifiers classifier NOUN bracis-19048 30 25 . . PUNCT bracis-19048 31 1 these these DET bracis-19048 31 2 clustering cluster VERB bracis-19048 31 3 measures measure NOUN bracis-19048 31 4 are be AUX bracis-19048 31 5 based base VERB bracis-19048 31 6 on on ADP bracis-19048 31 7 internal internal ADJ bracis-19048 31 8 indices index NOUN bracis-19048 31 9 which which PRON bracis-19048 31 10 extract extract VERB bracis-19048 31 11 information information NOUN bracis-19048 31 12 , , PUNCT bracis-19048 31 13 such such ADJ bracis-19048 31 14 as as ADP bracis-19048 31 15 compactness compactness NOUN bracis-19048 31 16 or or CCONJ bracis-19048 31 17 separation separation NOUN bracis-19048 31 18 , , PUNCT bracis-19048 31 19 to to PART bracis-19048 31 20 evaluate evaluate VERB bracis-19048 31 21 the the DET bracis-19048 31 22 goodness goodness NOUN bracis-19048 31 23 of of ADP bracis-19048 31 24 a a DET bracis-19048 31 25 clustering clustering ADJ bracis-19048 31 26 structure structure NOUN bracis-19048 31 27 . . PUNCT bracis-19048 32 1 although although SCONJ bracis-19048 32 2 some some PRON bracis-19048 32 3 of of ADP bracis-19048 32 4 these these DET bracis-19048 32 5 measures measure NOUN bracis-19048 32 6 have have AUX bracis-19048 32 7 been be AUX bracis-19048 32 8 used use VERB bracis-19048 32 9 in in ADP bracis-19048 32 10 unsupervised unsupervised ADJ bracis-19048 32 11 mtl mtl PROPN bracis-19048 32 12 scenarios scenario NOUN bracis-19048 32 13     SPACE bracis-19048 33 1 [ [ X bracis-19048 33 2 26 26 NUM bracis-19048 33 3 , , PUNCT bracis-19048 33 4 37 37 NUM bracis-19048 33 5 ] ] PUNCT bracis-19048 33 6 , , PUNCT bracis-19048 33 7 this this DET bracis-19048 33 8 work work NOUN bracis-19048 33 9 proposes propose VERB bracis-19048 33 10 the the DET bracis-19048 33 11 use use NOUN bracis-19048 33 12 of of ADP bracis-19048 33 13 those those DET bracis-19048 33 14 measures measure NOUN bracis-19048 33 15 on on ADP bracis-19048 33 16 classification classification NOUN bracis-19048 33 17 problems problem NOUN bracis-19048 33 18 . . PUNCT bracis-19048 34 1 the the DET bracis-19048 34 2 proposed propose VERB bracis-19048 34 3 approach approach NOUN bracis-19048 34 4 uses use VERB bracis-19048 34 5 class class NOUN bracis-19048 34 6 labels label NOUN bracis-19048 34 7 as as ADP bracis-19048 34 8 cluster cluster NOUN bracis-19048 34 9 indicators indicator NOUN bracis-19048 34 10 rather rather ADV bracis-19048 34 11 than than ADP bracis-19048 34 12 the the DET bracis-19048 34 13 numerical numerical ADJ bracis-19048 34 14 results result NOUN bracis-19048 34 15 of of ADP bracis-19048 34 16 a a DET bracis-19048 34 17 clustering cluster VERB bracis-19048 34 18 algorithm algorithm NOUN bracis-19048 34 19 , , PUNCT bracis-19048 34 20 and and CCONJ bracis-19048 34 21 is be AUX bracis-19048 34 22 expected expect VERB bracis-19048 34 23 to to PART bracis-19048 34 24 extract extract VERB bracis-19048 34 25 informative informative ADJ bracis-19048 34 26 measures measure NOUN bracis-19048 34 27 for for ADP bracis-19048 34 28 quantifying quantify VERB bracis-19048 34 29 statistical statistical ADJ bracis-19048 34 30 or or CCONJ bracis-19048 34 31 geometrical geometrical ADJ bracis-19048 34 32 characteristics characteristic NOUN bracis-19048 34 33 of of ADP bracis-19048 34 34 datasets dataset NOUN bracis-19048 34 35 . . PUNCT bracis-19048 35 1 the the DET bracis-19048 35 2 main main ADJ bracis-19048 35 3 goal goal NOUN bracis-19048 35 4 of of ADP bracis-19048 35 5 this this DET bracis-19048 35 6 paper paper NOUN bracis-19048 35 7 is be AUX bracis-19048 35 8 to to PART bracis-19048 35 9 investigate investigate VERB bracis-19048 35 10 whether whether SCONJ bracis-19048 35 11 clustering cluster VERB bracis-19048 35 12 measures measure NOUN bracis-19048 35 13 can can AUX bracis-19048 35 14 be be AUX bracis-19048 35 15 applied apply VERB bracis-19048 35 16 in in ADP bracis-19048 35 17 this this DET bracis-19048 35 18 context context NOUN bracis-19048 35 19 , , PUNCT bracis-19048 35 20 and and CCONJ bracis-19048 35 21 whether whether SCONJ bracis-19048 35 22 they they PRON bracis-19048 35 23 influence influence VERB bracis-19048 35 24 the the DET bracis-19048 35 25 choices choice NOUN bracis-19048 35 26 of of ADP bracis-19048 35 27 the the DET bracis-19048 35 28 recommendation recommendation NOUN bracis-19048 35 29 system system NOUN bracis-19048 35 30 . . PUNCT bracis-19048 36 1 the the DET bracis-19048 36 2 experimental experimental ADJ bracis-19048 36 3 results result NOUN bracis-19048 36 4 suggest suggest VERB bracis-19048 36 5 that that SCONJ bracis-19048 36 6 including include VERB bracis-19048 36 7 these these DET bracis-19048 36 8 measures measure NOUN bracis-19048 36 9 as as SCONJ bracis-19048 36 10 meta meta NOUN bracis-19048 36 11 - - PUNCT bracis-19048 36 12 features feature NOUN bracis-19048 36 13 contributes contribute VERB bracis-19048 36 14 to to ADP bracis-19048 36 15 more more ADV bracis-19048 36 16 accurate accurate ADJ bracis-19048 36 17 recommendations recommendation NOUN bracis-19048 36 18 in in ADP bracis-19048 36 19 the the DET bracis-19048 36 20 problem problem NOUN bracis-19048 36 21 of of ADP bracis-19048 36 22 classifier classifier NOUN bracis-19048 36 23 selection selection NOUN bracis-19048 36 24 . . PUNCT bracis-19048 37 1 additionally additionally ADV bracis-19048 37 2 , , PUNCT bracis-19048 37 3 an an DET bracis-19048 37 4 initial initial ADJ bracis-19048 37 5 evaluation evaluation NOUN bracis-19048 37 6 of of ADP bracis-19048 37 7 computational computational ADJ bracis-19048 37 8 costs cost NOUN bracis-19048 37 9 indicates indicate VERB bracis-19048 37 10 that that SCONJ bracis-19048 37 11 this this DET bracis-19048 37 12 mtl mtl PROPN bracis-19048 37 13 approach approach NOUN bracis-19048 37 14 is be AUX bracis-19048 37 15 substantially substantially ADV bracis-19048 37 16 less less ADV bracis-19048 37 17 computationally computationally ADV bracis-19048 37 18 expensive expensive ADJ bracis-19048 37 19 than than ADP bracis-19048 37 20 testing test VERB bracis-19048 37 21 all all DET bracis-19048 37 22 classifiers classifier NOUN bracis-19048 37 23 on on ADP bracis-19048 37 24 the the DET bracis-19048 37 25 dataset dataset NOUN bracis-19048 37 26 for for ADP bracis-19048 37 27 the the DET bracis-19048 37 28 selection selection NOUN bracis-19048 37 29 of of ADP bracis-19048 37 30 the the DET bracis-19048 37 31 best good ADJ bracis-19048 37 32 one one NUM bracis-19048 37 33 . . NOUN bracis-19048 37 34 2 2 NUM bracis-19048 37 35 background background NOUN bracis-19048 37 36 this this DET bracis-19048 37 37 section section NOUN bracis-19048 37 38 presents present VERB bracis-19048 37 39 the the DET bracis-19048 37 40 background background NOUN bracis-19048 37 41 information information NOUN bracis-19048 37 42 necessary necessary ADJ bracis-19048 37 43 to to PART bracis-19048 37 44 describe describe VERB bracis-19048 37 45 the the DET bracis-19048 37 46 proposed propose VERB bracis-19048 37 47 approach approach NOUN bracis-19048 37 48 : : PUNCT bracis-19048 37 49 sect sect NOUN bracis-19048 37 50 . . PUNCT bracis-19048 37 51     SPACE bracis-19048 37 52 2.1 2.1 NUM bracis-19048 37 53 introduces introduce NOUN bracis-19048 37 54 the the DET bracis-19048 37 55 meta meta ADV bracis-19048 37 56 - - PUNCT bracis-19048 37 57 learning learn VERB bracis-19048 37 58 framework framework NOUN bracis-19048 37 59 , , PUNCT bracis-19048 37 60 including include VERB bracis-19048 37 61 the the DET bracis-19048 37 62 process process NOUN bracis-19048 37 63 of of ADP bracis-19048 37 64 building build VERB bracis-19048 37 65 a a DET bracis-19048 37 66 meta meta ADV bracis-19048 37 67 - - PUNCT bracis-19048 37 68 dataset dataset NOUN bracis-19048 37 69 and and CCONJ bracis-19048 37 70 how how SCONJ bracis-19048 37 71 to to PART bracis-19048 37 72 recommend recommend VERB bracis-19048 37 73 algorithms algorithm NOUN bracis-19048 37 74 . . PUNCT bracis-19048 38 1 section section NOUN bracis-19048 38 2     SPACE bracis-19048 38 3 2.2 2.2 NUM bracis-19048 38 4 elaborates elaborate NOUN bracis-19048 38 5 on on ADP bracis-19048 38 6 clustering cluster VERB bracis-19048 38 7 meta meta NOUN bracis-19048 38 8 - - PUNCT bracis-19048 38 9 features feature NOUN bracis-19048 38 10 , , PUNCT bracis-19048 38 11 a a DET bracis-19048 38 12 subset subset NOUN bracis-19048 38 13 of of ADP bracis-19048 38 14 internal internal ADJ bracis-19048 38 15 indices index NOUN bracis-19048 38 16 based base VERB bracis-19048 38 17 on on ADP bracis-19048 38 18 compactness compactness NOUN bracis-19048 38 19 and and CCONJ bracis-19048 38 20 separation separation NOUN bracis-19048 38 21 of of ADP bracis-19048 38 22 the the DET bracis-19048 38 23 goodness goodness NOUN bracis-19048 38 24 of of ADP bracis-19048 38 25 a a DET bracis-19048 38 26 clustering clustering ADJ bracis-19048 38 27 structure structure NOUN bracis-19048 38 28 . . PUNCT bracis-19048 39 1 2.1 2.1 NUM bracis-19048 39 2 meta meta ADV bracis-19048 39 3 - - PUNCT bracis-19048 39 4 learning learn VERB bracis-19048 39 5 different different ADJ bracis-19048 39 6 algorithms algorithm NOUN bracis-19048 39 7 have have VERB bracis-19048 39 8 distinct distinct ADJ bracis-19048 39 9 learning learning NOUN bracis-19048 39 10 strategies strategy NOUN bracis-19048 39 11 , , PUNCT bracis-19048 39 12 thus thus ADV bracis-19048 39 13 the the DET bracis-19048 39 14 essence essence NOUN bracis-19048 39 15 of of ADP bracis-19048 39 16 learning learning NOUN bracis-19048 39 17 can can AUX bracis-19048 39 18 only only ADV bracis-19048 39 19 be be AUX bracis-19048 39 20 captured capture VERB bracis-19048 39 21 by by ADP bracis-19048 39 22 considering consider VERB bracis-19048 39 23 different different ADJ bracis-19048 39 24 learning learning NOUN bracis-19048 39 25 algorithms algorithm NOUN bracis-19048 39 26 with with ADP bracis-19048 39 27 diverse diverse ADJ bracis-19048 39 28 biases bias NOUN bracis-19048 39 29 to to PART bracis-19048 39 30 acquire acquire VERB bracis-19048 39 31 domain domain NOUN bracis-19048 39 32 specific specific ADJ bracis-19048 39 33 information information NOUN bracis-19048 39 34     SPACE bracis-19048 40 1 [ [ X bracis-19048 40 2 4 4 NUM bracis-19048 40 3 ] ] PUNCT bracis-19048 40 4 . . PUNCT bracis-19048 41 1 this this DET bracis-19048 41 2 concept concept NOUN bracis-19048 41 3 was be AUX bracis-19048 41 4 initially initially ADV bracis-19048 41 5 introduced introduce VERB bracis-19048 41 6 to to PART bracis-19048 41 7 make make VERB bracis-19048 41 8 the the DET bracis-19048 41 9 algorithm algorithm NOUN bracis-19048 41 10 selection selection NOUN bracis-19048 41 11 problem problem NOUN bracis-19048 41 12 systematic systematic NOUN bracis-19048 41 13     SPACE bracis-19048 42 1 [ [ X bracis-19048 42 2 30 30 NUM bracis-19048 42 3 ] ] PUNCT bracis-19048 42 4 , , PUNCT bracis-19048 42 5 and and CCONJ bracis-19048 42 6 the the DET bracis-19048 42 7 goal goal NOUN bracis-19048 42 8 is be AUX bracis-19048 42 9 to to PART bracis-19048 42 10 predict predict VERB bracis-19048 42 11 the the DET bracis-19048 42 12 best good ADJ bracis-19048 42 13 algorithm algorithm NOUN bracis-19048 42 14 to to PART bracis-19048 42 15 solve solve VERB bracis-19048 42 16 a a DET bracis-19048 42 17 specific specific ADJ bracis-19048 42 18 given give VERB bracis-19048 42 19 problem problem NOUN bracis-19048 42 20 , , PUNCT bracis-19048 42 21 when when SCONJ bracis-19048 42 22 more more ADJ bracis-19048 42 23 than than ADP bracis-19048 42 24 one one NUM bracis-19048 42 25 option option NOUN bracis-19048 42 26 is be AUX bracis-19048 42 27 available available ADJ bracis-19048 42 28 . . PUNCT bracis-19048 43 1 the the DET bracis-19048 43 2 components component NOUN bracis-19048 43 3 of of ADP bracis-19048 43 4 this this DET bracis-19048 43 5 model model NOUN bracis-19048 43 6 are be AUX bracis-19048 43 7 the the DET bracis-19048 43 8 space space NOUN bracis-19048 43 9 of of ADP bracis-19048 43 10 problem problem NOUN bracis-19048 43 11 instances instance NOUN bracis-19048 43 12 ( ( PUNCT bracis-19048 43 13 \(\mathcal \(\mathcal ADJ bracis-19048 43 14 { { PUNCT bracis-19048 43 15 p}\ p}\ NUM bracis-19048 43 16 ) ) PUNCT bracis-19048 43 17 ) ) PUNCT bracis-19048 43 18 ; ; PUNCT bracis-19048 43 19 the the DET bracis-19048 43 20 space space NOUN bracis-19048 43 21 of of ADP bracis-19048 43 22 instance instance NOUN bracis-19048 43 23 meta meta X bracis-19048 43 24 - - PUNCT bracis-19048 43 25 features feature NOUN bracis-19048 43 26 ( ( PUNCT bracis-19048 43 27 \(\mathcal \(\mathcal ADJ bracis-19048 43 28 { { PUNCT bracis-19048 43 29 f}\ f}\ NOUN bracis-19048 43 30 ) ) PUNCT bracis-19048 43 31 ) ) PUNCT bracis-19048 43 32 ; ; PUNCT bracis-19048 43 33 the the DET bracis-19048 43 34 space space NOUN bracis-19048 43 35 of of ADP bracis-19048 43 36 algorithms algorithm NOUN bracis-19048 43 37 ( ( PUNCT bracis-19048 43 38 \(\mathcal \(\mathcal ADJ bracis-19048 43 39 { { PUNCT bracis-19048 43 40 a}\ a}\ NUM bracis-19048 43 41 ) ) PUNCT bracis-19048 43 42 ) ) PUNCT bracis-19048 43 43 ; ; PUNCT bracis-19048 43 44 and and CCONJ bracis-19048 43 45 the the DET bracis-19048 43 46 space space NOUN bracis-19048 43 47 of of ADP bracis-19048 43 48 evaluation evaluation NOUN bracis-19048 43 49 measures measure NOUN bracis-19048 43 50 ( ( PUNCT bracis-19048 43 51 \(\mathcal \(\mathcal ADJ bracis-19048 43 52 { { PUNCT bracis-19048 43 53 y}\ y}\ NOUN bracis-19048 43 54 ) ) PUNCT bracis-19048 43 55 ) ) PUNCT bracis-19048 43 56 . . PUNCT bracis-19048 44 1 from from ADP bracis-19048 44 2 these these DET bracis-19048 44 3 components component NOUN bracis-19048 44 4 , , PUNCT bracis-19048 44 5 an an DET bracis-19048 44 6 mtl mtl PROPN bracis-19048 44 7 system system NOUN bracis-19048 44 8 can can AUX bracis-19048 44 9 obtain obtain VERB bracis-19048 44 10 a a DET bracis-19048 44 11 model model NOUN bracis-19048 44 12 capable capable ADJ bracis-19048 44 13 of of ADP bracis-19048 44 14 mapping map VERB bracis-19048 44 15 a a DET bracis-19048 44 16 dataset dataset NOUN bracis-19048 44 17 or or CCONJ bracis-19048 44 18 problem problem NOUN bracis-19048 44 19 \(p\in \(p\in PUNCT bracis-19048 44 20 \mathcal \mathcal PROPN bracis-19048 44 21 { { PUNCT bracis-19048 44 22 p}\ p}\ NUM bracis-19048 44 23 ) ) PUNCT bracis-19048 44 24 , , PUNCT bracis-19048 44 25 described describe VERB bracis-19048 44 26 by by ADP bracis-19048 44 27 meta meta NOUN bracis-19048 44 28 - - PUNCT bracis-19048 44 29 features feature VERB bracis-19048 44 30 \(f\in \(f\in PUNCT bracis-19048 44 31 \mathcal \mathcal X bracis-19048 44 32 { { PUNCT bracis-19048 44 33 f}\ f}\ NOUN bracis-19048 44 34 ) ) PUNCT bracis-19048 44 35 , , PUNCT bracis-19048 44 36 into into ADP bracis-19048 44 37 one one NUM bracis-19048 44 38 algorithm algorithm NOUN bracis-19048 44 39 \(\alpha \(\alpha PART bracis-19048 44 40 \in \in PROPN bracis-19048 44 41 \mathcal \mathcal PROPN bracis-19048 44 42 { { PUNCT bracis-19048 44 43 a}\ a}\ NUM bracis-19048 44 44 ) ) PUNCT bracis-19048 44 45 able able ADJ bracis-19048 44 46 to to PART bracis-19048 44 47 solve solve VERB bracis-19048 44 48 the the DET bracis-19048 44 49 problem problem NOUN bracis-19048 44 50 with with ADP bracis-19048 44 51 a a DET bracis-19048 44 52 good good ADJ bracis-19048 44 53 predictive predictive ADJ bracis-19048 44 54 performance performance NOUN bracis-19048 44 55 according accord VERB bracis-19048 44 56 to to ADP bracis-19048 44 57 measure measure NOUN bracis-19048 44 58 \(y\in \(y\in PUNCT bracis-19048 44 59 \mathcal \mathcal PROPN bracis-19048 44 60 { { PUNCT bracis-19048 44 61 y}\ y}\ NOUN bracis-19048 44 62 ) ) PUNCT bracis-19048 44 63 . . PUNCT bracis-19048 45 1 the the DET bracis-19048 45 2 meta meta ADJ bracis-19048 45 3 - - PUNCT bracis-19048 45 4 learner learner NOUN bracis-19048 45 5 recommendation recommendation NOUN bracis-19048 45 6 would would AUX bracis-19048 45 7 be be AUX bracis-19048 45 8 the the DET bracis-19048 45 9 algorithm algorithm NOUN bracis-19048 45 10 with with ADP bracis-19048 45 11 the the DET bracis-19048 45 12 best good ADJ bracis-19048 45 13 expected expect VERB bracis-19048 45 14 \(y(\alpha \(y(\alpha X bracis-19048 45 15 ( ( PUNCT bracis-19048 45 16 p))\ p))\ PROPN bracis-19048 45 17 ) ) PUNCT bracis-19048 45 18 . . PUNCT bracis-19048 46 1 this this PRON bracis-19048 46 2 can can AUX bracis-19048 46 3 be be AUX bracis-19048 46 4 further far ADV bracis-19048 46 5 improved improve VERB bracis-19048 46 6 , , PUNCT bracis-19048 46 7 for for ADP bracis-19048 46 8 instance instance NOUN bracis-19048 46 9 , , PUNCT bracis-19048 46 10 by by ADP bracis-19048 46 11 the the DET bracis-19048 46 12 inclusion inclusion NOUN bracis-19048 46 13 of of ADP bracis-19048 46 14 components component NOUN bracis-19048 46 15 which which PRON bracis-19048 46 16 may may AUX bracis-19048 46 17 guide guide VERB bracis-19048 46 18 theoretical theoretical ADJ bracis-19048 46 19 support support NOUN bracis-19048 46 20 to to PART bracis-19048 46 21 refine refine VERB bracis-19048 46 22 the the DET bracis-19048 46 23 recommendation recommendation NOUN bracis-19048 46 24 system system NOUN bracis-19048 46 25     SPACE bracis-19048 47 1 [ [ X bracis-19048 47 2 34 34 NUM bracis-19048 47 3 ] ] PUNCT bracis-19048 47 4 . . PUNCT bracis-19048 48 1 a a DET bracis-19048 48 2 crucial crucial ADJ bracis-19048 48 3 component component NOUN bracis-19048 48 4 of of ADP bracis-19048 48 5 these these DET bracis-19048 48 6 previous previous ADJ bracis-19048 48 7 approaches approach NOUN bracis-19048 48 8 is be AUX bracis-19048 48 9 the the DET bracis-19048 48 10 definition definition NOUN bracis-19048 48 11 of of ADP bracis-19048 48 12 the the DET bracis-19048 48 13 meta meta NOUN bracis-19048 48 14 - - PUNCT bracis-19048 48 15 features feature NOUN bracis-19048 48 16 ( ( PUNCT bracis-19048 48 17 \(\mathcal \(\mathcal ADJ bracis-19048 48 18 { { PUNCT bracis-19048 48 19 f}\ f}\ NOUN bracis-19048 48 20 ) ) PUNCT bracis-19048 48 21 ) ) PUNCT bracis-19048 48 22 used use VERB bracis-19048 48 23 to to PART bracis-19048 48 24 describe describe VERB bracis-19048 48 25 general general ADJ bracis-19048 48 26 properties property NOUN bracis-19048 48 27 of of ADP bracis-19048 48 28 datasets dataset NOUN bracis-19048 48 29     SPACE bracis-19048 49 1 [ [ X bracis-19048 49 2 3 3 NUM bracis-19048 49 3 ] ] PUNCT bracis-19048 49 4 . . PUNCT bracis-19048 50 1 they they PRON bracis-19048 50 2 must must AUX bracis-19048 50 3 be be AUX bracis-19048 50 4 able able ADJ bracis-19048 50 5 to to PART bracis-19048 50 6 provide provide VERB bracis-19048 50 7 evidence evidence NOUN bracis-19048 50 8 on on ADP bracis-19048 50 9 the the DET bracis-19048 50 10 future future ADJ bracis-19048 50 11 performance performance NOUN bracis-19048 50 12 of of ADP bracis-19048 50 13 the the DET bracis-19048 50 14 algorithms algorithm NOUN bracis-19048 50 15 in in ADP bracis-19048 50 16 \(\mathcal \(\mathcal ADJ bracis-19048 50 17 { { PUNCT bracis-19048 50 18 a}\ a}\ NUM bracis-19048 50 19 ) ) PUNCT bracis-19048 50 20 and and CCONJ bracis-19048 50 21 to to PART bracis-19048 50 22 discriminate discriminate VERB bracis-19048 50 23 , , PUNCT bracis-19048 50 24 with with ADP bracis-19048 50 25 an an DET bracis-19048 50 26 acceptable acceptable ADJ bracis-19048 50 27 computational computational ADJ bracis-19048 50 28 cost cost NOUN bracis-19048 50 29 , , PUNCT bracis-19048 50 30 the the DET bracis-19048 50 31 performance performance NOUN bracis-19048 50 32 of of ADP bracis-19048 50 33 a a DET bracis-19048 50 34 group group NOUN bracis-19048 50 35 of of ADP bracis-19048 50 36 algorithms algorithm NOUN bracis-19048 50 37 . . PUNCT bracis-19048 51 1 the the DET bracis-19048 51 2 main main ADJ bracis-19048 51 3 meta meta ADJ bracis-19048 51 4 - - PUNCT bracis-19048 51 5 features feature NOUN bracis-19048 51 6 used use VERB bracis-19048 51 7 in in ADP bracis-19048 51 8 the the DET bracis-19048 51 9 mtl mtl PROPN bracis-19048 51 10 literature literature NOUN bracis-19048 51 11 can can AUX bracis-19048 51 12 be be AUX bracis-19048 51 13 divided divide VERB bracis-19048 51 14 into into ADP bracis-19048 51 15 six six NUM bracis-19048 51 16 main main ADJ bracis-19048 51 17 groups group NOUN bracis-19048 51 18     SPACE bracis-19048 52 1 [ [ X bracis-19048 52 2 31 31 NUM bracis-19048 52 3 ] ] PUNCT bracis-19048 52 4 , , PUNCT bracis-19048 52 5 called call VERB bracis-19048 52 6 standard standard ADJ bracis-19048 52 7 meta meta NOUN bracis-19048 52 8 - - PUNCT bracis-19048 52 9 features feature NOUN bracis-19048 52 10 in in ADP bracis-19048 52 11 this this DET bracis-19048 52 12 work work NOUN bracis-19048 52 13 . . PUNCT bracis-19048 53 1 they they PRON bracis-19048 53 2 represent represent VERB bracis-19048 53 3 meta meta NOUN bracis-19048 53 4 - - PUNCT bracis-19048 53 5 features feature NOUN bracis-19048 53 6 based base VERB bracis-19048 53 7 general general ADJ bracis-19048 53 8 high high ADJ bracis-19048 53 9 - - PUNCT bracis-19048 53 10 level level NOUN bracis-19048 53 11 summaries summary NOUN bracis-19048 53 12 of of ADP bracis-19048 53 13 the the DET bracis-19048 53 14 data datum NOUN bracis-19048 53 15 , , PUNCT bracis-19048 53 16 statistical statistical ADJ bracis-19048 53 17 and and CCONJ bracis-19048 53 18 information information NOUN bracis-19048 53 19 theory theory NOUN bracis-19048 53 20 properties property NOUN bracis-19048 53 21 of of ADP bracis-19048 53 22 the the DET bracis-19048 53 23 data datum NOUN bracis-19048 53 24 , , PUNCT bracis-19048 53 25 properties property NOUN bracis-19048 53 26 of of ADP bracis-19048 53 27 decision decision NOUN bracis-19048 53 28 trees tree NOUN bracis-19048 53 29 ( ( PUNCT bracis-19048 53 30 dts dts NOUN bracis-19048 53 31 ) ) PUNCT bracis-19048 53 32 induced induce VERB bracis-19048 53 33 from from ADP bracis-19048 53 34 the the DET bracis-19048 53 35 data datum NOUN bracis-19048 53 36 and and CCONJ bracis-19048 53 37 the the DET bracis-19048 53 38 performance performance NOUN bracis-19048 53 39 of of ADP bracis-19048 53 40 simple simple ADJ bracis-19048 53 41 and and CCONJ bracis-19048 53 42 fast fast ADJ bracis-19048 53 43 learning learn VERB bracis-19048 53 44 algorithms algorithm NOUN bracis-19048 53 45     SPACE bracis-19048 54 1 [ [ X bracis-19048 54 2 9 9 NUM bracis-19048 54 3 , , PUNCT bracis-19048 54 4 25 25 NUM bracis-19048 54 5 , , PUNCT bracis-19048 54 6 29 29 NUM bracis-19048 54 7 ] ] PUNCT bracis-19048 54 8 . . PUNCT bracis-19048 55 1 another another DET bracis-19048 55 2 concern concern NOUN bracis-19048 55 3 is be AUX bracis-19048 55 4 the the DET bracis-19048 55 5 definition definition NOUN bracis-19048 55 6 of of ADP bracis-19048 55 7 the the DET bracis-19048 55 8 set set NOUN bracis-19048 55 9 of of ADP bracis-19048 55 10 problem problem NOUN bracis-19048 55 11 instances instance NOUN bracis-19048 55 12 ( ( PUNCT bracis-19048 55 13 \(\mathcal \(\mathcal ADJ bracis-19048 55 14 { { PUNCT bracis-19048 55 15 p}\ p}\ NUM bracis-19048 55 16 ) ) PUNCT bracis-19048 55 17 ) ) PUNCT bracis-19048 55 18 , , PUNCT bracis-19048 55 19 since since SCONJ bracis-19048 55 20 the the DET bracis-19048 55 21 use use NOUN bracis-19048 55 22 of of ADP bracis-19048 55 23 a a DET bracis-19048 55 24 large large ADJ bracis-19048 55 25 number number NOUN bracis-19048 55 26 of of ADP bracis-19048 55 27 diverse diverse ADJ bracis-19048 55 28 datasets dataset NOUN bracis-19048 55 29 is be AUX bracis-19048 55 30 recommended recommend VERB bracis-19048 55 31 to to PART bracis-19048 55 32 induce induce VERB bracis-19048 55 33 a a DET bracis-19048 55 34 reliable reliable ADJ bracis-19048 55 35 meta meta ADJ bracis-19048 55 36 - - PUNCT bracis-19048 55 37 model model NOUN bracis-19048 55 38     SPACE bracis-19048 56 1 [ [ X bracis-19048 56 2 4 4 NUM bracis-19048 56 3 ] ] PUNCT bracis-19048 56 4 . . PUNCT bracis-19048 57 1 attempts attempt NOUN bracis-19048 57 2 to to PART bracis-19048 57 3 reduce reduce VERB bracis-19048 57 4 the the DET bracis-19048 57 5 bias bias NOUN bracis-19048 57 6 in in ADP bracis-19048 57 7 this this DET bracis-19048 57 8 choice choice NOUN bracis-19048 57 9 include include VERB bracis-19048 57 10 using use VERB bracis-19048 57 11 datasets dataset NOUN bracis-19048 57 12 from from ADP bracis-19048 57 13 different different ADJ bracis-19048 57 14 data datum NOUN bracis-19048 57 15 repositories repository NOUN bracis-19048 57 16 , , PUNCT bracis-19048 57 17 such such ADJ bracis-19048 57 18 as as ADP bracis-19048 57 19 uci uci PROPN bracis-19048 57 20     SPACE bracis-19048 58 1 [ [ X bracis-19048 58 2 13 13 NUM bracis-19048 58 3 ] ] PUNCT bracis-19048 58 4 and and CCONJ bracis-19048 58 5 openml openml NOUN bracis-19048 58 6     SPACE bracis-19048 59 1 [ [ X bracis-19048 59 2 36 36 NUM bracis-19048 59 3 ] ] PUNCT bracis-19048 59 4 . . PUNCT bracis-19048 60 1 the the DET bracis-19048 60 2 importance importance NOUN bracis-19048 60 3 of of ADP bracis-19048 60 4 problem problem NOUN bracis-19048 60 5 diversity diversity NOUN bracis-19048 60 6 is be AUX bracis-19048 60 7 based base VERB bracis-19048 60 8 on on ADP bracis-19048 60 9 the the DET bracis-19048 60 10 underlying underlying ADJ bracis-19048 60 11 assumption assumption NOUN bracis-19048 60 12 that that SCONJ bracis-19048 60 13 the the DET bracis-19048 60 14 meta meta NOUN bracis-19048 60 15 - - PUNCT bracis-19048 60 16 model model NOUN bracis-19048 60 17 is be AUX bracis-19048 60 18 expected expect VERB bracis-19048 60 19 to to PART bracis-19048 60 20 generalize generalize VERB bracis-19048 60 21 the the DET bracis-19048 60 22 acquired acquire VERB bracis-19048 60 23 knowledge knowledge NOUN bracis-19048 60 24 when when SCONJ bracis-19048 60 25 faced face VERB bracis-19048 60 26 with with ADP bracis-19048 60 27 new new ADJ bracis-19048 60 28 problem problem NOUN bracis-19048 60 29 instances instance NOUN bracis-19048 60 30 without without ADP bracis-19048 60 31 explicit explicit ADJ bracis-19048 60 32 constraints constraint NOUN bracis-19048 60 33 in in ADP bracis-19048 60 34 terms term NOUN bracis-19048 60 35 of of ADP bracis-19048 60 36 expected expect VERB bracis-19048 60 37 problem problem NOUN bracis-19048 60 38 characteristics characteristic NOUN bracis-19048 60 39 . . PUNCT bracis-19048 61 1 the the DET bracis-19048 61 2 selected select VERB bracis-19048 61 3 algorithms algorithms NOUN bracis-19048 61 4 \(\mathcal \(\mathcal ADJ bracis-19048 61 5 { { PUNCT bracis-19048 61 6 a}\ a}\ NUM bracis-19048 61 7 ) ) PUNCT bracis-19048 61 8 represent represent VERB bracis-19048 61 9 a a DET bracis-19048 61 10 set set NOUN bracis-19048 61 11 of of ADP bracis-19048 61 12 candidate candidate NOUN bracis-19048 61 13 algorithms algorithm NOUN bracis-19048 61 14 to to PART bracis-19048 61 15 be be AUX bracis-19048 61 16 recommended recommend VERB bracis-19048 61 17 in in ADP bracis-19048 61 18 the the DET bracis-19048 61 19 algorithm algorithm NOUN bracis-19048 61 20 selection selection NOUN bracis-19048 61 21 process process NOUN bracis-19048 61 22 . . PUNCT bracis-19048 62 1 ideally ideally ADV bracis-19048 62 2 , , PUNCT bracis-19048 62 3 these these DET bracis-19048 62 4 algorithms algorithm NOUN bracis-19048 62 5 should should AUX bracis-19048 62 6 also also ADV bracis-19048 62 7 be be AUX bracis-19048 62 8 sufficiently sufficiently ADV bracis-19048 62 9 different different ADJ bracis-19048 62 10 from from ADP bracis-19048 62 11 each each DET bracis-19048 62 12 other other ADJ bracis-19048 62 13 and and CCONJ bracis-19048 62 14 represent represent VERB bracis-19048 62 15 all all DET bracis-19048 62 16 regions region NOUN bracis-19048 62 17 in in ADP bracis-19048 62 18 the the DET bracis-19048 62 19 algorithm algorithm NOUN bracis-19048 62 20 space space NOUN bracis-19048 63 1 [ [ X bracis-19048 63 2 34 34 NUM bracis-19048 63 3 ] ] PUNCT bracis-19048 63 4 . . PUNCT bracis-19048 64 1 the the DET bracis-19048 64 2 models model NOUN bracis-19048 64 3 induced induce VERB bracis-19048 64 4 by by ADP bracis-19048 64 5 the the DET bracis-19048 64 6 algorithms algorithm NOUN bracis-19048 64 7 can can AUX bracis-19048 64 8 be be AUX bracis-19048 64 9 evaluated evaluate VERB bracis-19048 64 10 by by ADP bracis-19048 64 11 different different ADJ bracis-19048 64 12 measures measure NOUN bracis-19048 64 13 . . PUNCT bracis-19048 65 1 quality quality NOUN bracis-19048 65 2 measures measure NOUN bracis-19048 65 3 \(\mathcal \(\mathcal ADJ bracis-19048 65 4 { { PUNCT bracis-19048 65 5 y}\ y}\ NOUN bracis-19048 65 6 ) ) PUNCT bracis-19048 65 7 for for ADP bracis-19048 65 8 assessing assess VERB bracis-19048 65 9 the the DET bracis-19048 65 10 models model NOUN bracis-19048 65 11 depend depend VERB bracis-19048 65 12 on on ADP bracis-19048 65 13 the the DET bracis-19048 65 14 nature nature NOUN bracis-19048 65 15 of of ADP bracis-19048 65 16 the the DET bracis-19048 65 17 problem problem NOUN bracis-19048 65 18 . . PUNCT bracis-19048 66 1 classifiers classifier NOUN bracis-19048 66 2 , , PUNCT bracis-19048 66 3 for for ADP bracis-19048 66 4 example example NOUN bracis-19048 66 5 , , PUNCT bracis-19048 66 6 can can AUX bracis-19048 66 7 be be AUX bracis-19048 66 8 evaluated evaluate VERB bracis-19048 66 9 by by ADP bracis-19048 66 10 different different ADJ bracis-19048 66 11 measures measure NOUN bracis-19048 66 12 such such ADJ bracis-19048 66 13 as as ADP bracis-19048 66 14 accuracy accuracy NOUN bracis-19048 66 15 , , PUNCT bracis-19048 66 16 \(f_\beta \(f_\beta PROPN bracis-19048 66 17 \ \ PROPN bracis-19048 66 18 ) ) PUNCT bracis-19048 66 19 , , PUNCT bracis-19048 66 20 area area NOUN bracis-19048 66 21 under under ADP bracis-19048 66 22 the the DET bracis-19048 66 23 roc roc PROPN bracis-19048 66 24 curve curve NOUN bracis-19048 66 25 ( ( PUNCT bracis-19048 66 26 auc auc NOUN bracis-19048 66 27 ) ) PUNCT bracis-19048 66 28 or or CCONJ bracis-19048 66 29 the the DET bracis-19048 66 30 kappa kappa PROPN bracis-19048 66 31 coefficient coefficient PROPN bracis-19048 66 32 , , PUNCT bracis-19048 66 33 among among ADP bracis-19048 66 34 others other NOUN bracis-19048 66 35 . . PUNCT bracis-19048 67 1 the the DET bracis-19048 67 2 step step NOUN bracis-19048 67 3 following follow VERB bracis-19048 67 4 the the DET bracis-19048 67 5 extraction extraction NOUN bracis-19048 67 6 of of ADP bracis-19048 67 7 meta meta NOUN bracis-19048 67 8 - - PUNCT bracis-19048 67 9 features feature NOUN bracis-19048 67 10 from from ADP bracis-19048 67 11 the the DET bracis-19048 67 12 datasets dataset NOUN bracis-19048 67 13 and and CCONJ bracis-19048 67 14 training train VERB bracis-19048 67 15 the the DET bracis-19048 67 16 set set NOUN bracis-19048 67 17 of of ADP bracis-19048 67 18 algorithms algorithm NOUN bracis-19048 67 19 is be AUX bracis-19048 67 20 the the DET bracis-19048 67 21 labeling labeling NOUN bracis-19048 67 22 of of ADP bracis-19048 67 23 meta meta NOUN bracis-19048 67 24 - - PUNCT bracis-19048 67 25 examples example NOUN bracis-19048 67 26 in in ADP bracis-19048 67 27 the the DET bracis-19048 67 28 meta meta ADJ bracis-19048 67 29 - - PUNCT bracis-19048 67 30 base base NOUN bracis-19048 67 31 . . PUNCT bracis-19048 68 1 the the DET bracis-19048 68 2 three three NUM bracis-19048 68 3 properties property NOUN bracis-19048 68 4 most most ADV bracis-19048 68 5 frequently frequently ADV bracis-19048 68 6 used use VERB bracis-19048 68 7 in in ADP bracis-19048 68 8 this this DET bracis-19048 68 9 task task NOUN bracis-19048 68 10 are be AUX bracis-19048 68 11     SPACE bracis-19048 68 12 [ [ X bracis-19048 68 13 4 4 NUM bracis-19048 68 14 ] ] PUNCT bracis-19048 68 15 : : PUNCT bracis-19048 68 16 ( ( PUNCT bracis-19048 68 17 i i NOUN bracis-19048 68 18 ) ) PUNCT bracis-19048 68 19 the the DET bracis-19048 68 20 best well ADV bracis-19048 68 21 performing perform VERB bracis-19048 68 22 algorithm algorithm NOUN bracis-19048 68 23 on on ADP bracis-19048 68 24 the the DET bracis-19048 68 25 meta meta ADJ bracis-19048 68 26 - - PUNCT bracis-19048 68 27 example example NOUN bracis-19048 68 28 ’s ’s PART bracis-19048 68 29 dataset dataset NOUN bracis-19048 68 30 ( ( PUNCT bracis-19048 68 31 a a DET bracis-19048 68 32 meta meta ADJ bracis-19048 68 33 - - PUNCT bracis-19048 68 34 classification classification NOUN bracis-19048 68 35 problem problem NOUN bracis-19048 68 36 ) ) PUNCT bracis-19048 68 37 ; ; PUNCT bracis-19048 68 38 ( ( PUNCT bracis-19048 68 39 ii ii X bracis-19048 68 40 ) ) PUNCT bracis-19048 68 41 the the DET bracis-19048 68 42 ranking ranking NOUN bracis-19048 68 43 of of ADP bracis-19048 68 44 the the DET bracis-19048 68 45 algorithm algorithm NOUN bracis-19048 68 46 according accord VERB bracis-19048 68 47 to to ADP bracis-19048 68 48 its its PRON bracis-19048 68 49 performance performance NOUN bracis-19048 68 50 ( ( PUNCT bracis-19048 68 51 a a DET bracis-19048 68 52 meta meta ADV bracis-19048 68 53 - - PUNCT bracis-19048 68 54 ranking ranking ADJ bracis-19048 68 55 problem problem NOUN bracis-19048 68 56 ) ) PUNCT bracis-19048 68 57 ; ; PUNCT bracis-19048 68 58 and and CCONJ bracis-19048 68 59 ( ( PUNCT bracis-19048 68 60 iii iii X bracis-19048 68 61 ) ) PUNCT bracis-19048 68 62 the the DET bracis-19048 68 63 raw raw ADJ bracis-19048 68 64 performance performance NOUN bracis-19048 68 65 value value NOUN bracis-19048 68 66 of of ADP bracis-19048 68 67 each each DET bracis-19048 68 68 algorithm algorithm NOUN bracis-19048 68 69 on on ADP bracis-19048 68 70 the the DET bracis-19048 68 71 dataset dataset NOUN bracis-19048 68 72 ( ( PUNCT bracis-19048 68 73 a a DET bracis-19048 68 74 meta meta ADJ bracis-19048 68 75 - - PUNCT bracis-19048 68 76 regression regression NOUN bracis-19048 68 77 problem problem NOUN bracis-19048 68 78 ) ) PUNCT bracis-19048 68 79 . . PUNCT bracis-19048 69 1 differently differently ADV bracis-19048 69 2 from from ADP bracis-19048 69 3 previous previous ADJ bracis-19048 69 4 works work NOUN bracis-19048 69 5 on on ADP bracis-19048 69 6 mtl mtl PROPN bracis-19048 69 7 applied apply VERB bracis-19048 69 8 on on ADP bracis-19048 69 9 clustering cluster VERB bracis-19048 69 10 data datum NOUN bracis-19048 69 11     SPACE bracis-19048 70 1 [ [ X bracis-19048 70 2 26 26 NUM bracis-19048 70 3 , , PUNCT bracis-19048 70 4 37 37 NUM bracis-19048 70 5 ] ] PUNCT bracis-19048 70 6 , , PUNCT bracis-19048 70 7 this this DET bracis-19048 70 8 work work NOUN bracis-19048 70 9 presents present VERB bracis-19048 70 10 a a DET bracis-19048 70 11 mtl mtl PROPN bracis-19048 70 12 regression regression NOUN bracis-19048 70 13 task task NOUN bracis-19048 70 14 based base VERB bracis-19048 70 15 on on ADP bracis-19048 70 16 the the DET bracis-19048 70 17 use use NOUN bracis-19048 70 18 of of ADP bracis-19048 70 19 clustering cluster VERB bracis-19048 70 20 meta meta NOUN bracis-19048 70 21 - - PUNCT bracis-19048 70 22 features feature NOUN bracis-19048 70 23 using use VERB bracis-19048 70 24 internal internal ADJ bracis-19048 70 25 indices index NOUN bracis-19048 70 26 , , PUNCT bracis-19048 70 27 to to PART bracis-19048 70 28 extract extract VERB bracis-19048 70 29 information information NOUN bracis-19048 70 30 like like ADP bracis-19048 70 31 separation separation NOUN bracis-19048 70 32 and and CCONJ bracis-19048 70 33 compactness compactness NOUN bracis-19048 70 34 on on ADP bracis-19048 70 35 a a DET bracis-19048 70 36 supervised supervised ADJ bracis-19048 70 37 scenario scenario NOUN bracis-19048 70 38 using use VERB bracis-19048 70 39 the the DET bracis-19048 70 40 class class NOUN bracis-19048 70 41 labels label NOUN bracis-19048 70 42 . . PUNCT bracis-19048 71 1 in in ADP bracis-19048 71 2 this this DET bracis-19048 71 3 case case NOUN bracis-19048 71 4 , , PUNCT bracis-19048 71 5 the the DET bracis-19048 71 6 main main ADJ bracis-19048 71 7 objective objective NOUN bracis-19048 71 8 is be AUX bracis-19048 71 9 to to PART bracis-19048 71 10 improve improve VERB bracis-19048 71 11 the the DET bracis-19048 71 12 algorithm algorithm NOUN bracis-19048 71 13 recommendation recommendation NOUN bracis-19048 71 14 using use VERB bracis-19048 71 15 internal internal ADJ bracis-19048 71 16 indices index NOUN bracis-19048 71 17 , , PUNCT bracis-19048 71 18 motivated motivate VERB bracis-19048 71 19 by by ADP bracis-19048 71 20 their their PRON bracis-19048 71 21 generally generally ADV bracis-19048 71 22 low low ADJ bracis-19048 71 23 computational computational ADJ bracis-19048 71 24 cost cost NOUN bracis-19048 71 25 and and CCONJ bracis-19048 71 26 high high ADJ bracis-19048 71 27 degree degree NOUN bracis-19048 71 28 of of ADP bracis-19048 71 29 information information NOUN bracis-19048 71 30 . . PUNCT bracis-19048 72 1 the the DET bracis-19048 72 2 standard standard ADJ bracis-19048 72 3 meta meta NOUN bracis-19048 72 4 - - PUNCT bracis-19048 72 5 features feature VERB bracis-19048 72 6     SPACE bracis-19048 73 1 [ [ PUNCT bracis-19048 73 2 31 31 NUM bracis-19048 73 3 ] ] PUNCT bracis-19048 73 4 are be AUX bracis-19048 73 5 considered consider VERB bracis-19048 73 6 to to PART bracis-19048 73 7 provide provide VERB bracis-19048 73 8 a a DET bracis-19048 73 9 comparison comparison NOUN bracis-19048 73 10 baseline baseline NOUN bracis-19048 73 11 and and CCONJ bracis-19048 73 12 to to PART bracis-19048 73 13 allow allow VERB bracis-19048 73 14 an an DET bracis-19048 73 15 objective objective ADJ bracis-19048 73 16 analysis analysis NOUN bracis-19048 73 17 of of ADP bracis-19048 73 18 the the DET bracis-19048 73 19 variation variation NOUN bracis-19048 73 20 in in ADP bracis-19048 73 21 performance performance NOUN bracis-19048 73 22 resulting result VERB bracis-19048 73 23 from from ADP bracis-19048 73 24 using use VERB bracis-19048 73 25 the the DET bracis-19048 73 26 new new ADJ bracis-19048 73 27 clustering clustering ADJ bracis-19048 73 28 meta meta NOUN bracis-19048 73 29 - - PUNCT bracis-19048 73 30 features feature NOUN bracis-19048 73 31 . . PUNCT bracis-19048 74 1 the the DET bracis-19048 74 2 evaluation evaluation NOUN bracis-19048 74 3 of of ADP bracis-19048 74 4 the the DET bracis-19048 74 5 recommender recommender NOUN bracis-19048 74 6 system system NOUN bracis-19048 74 7 includes include VERB bracis-19048 74 8 assessing assess VERB bracis-19048 74 9 the the DET bracis-19048 74 10 performance performance NOUN bracis-19048 74 11 in in ADP bracis-19048 74 12 the the DET bracis-19048 74 13 baseand baseand NOUN bracis-19048 74 14 meta meta ADJ bracis-19048 74 15 - - PUNCT bracis-19048 74 16 levels level NOUN bracis-19048 74 17 and and CCONJ bracis-19048 74 18 the the DET bracis-19048 74 19 cost cost NOUN bracis-19048 74 20 in in ADP bracis-19048 74 21 execution execution NOUN bracis-19048 74 22 time time NOUN bracis-19048 74 23 . . PUNCT bracis-19048 75 1 2.2 2.2 NUM bracis-19048 75 2 clustering cluster VERB bracis-19048 75 3 meta meta NOUN bracis-19048 75 4 - - PUNCT bracis-19048 75 5 features feature VERB bracis-19048 75 6 a a DET bracis-19048 75 7 few few ADJ bracis-19048 75 8 definitions definition NOUN bracis-19048 75 9 must must AUX bracis-19048 75 10 be be AUX bracis-19048 75 11 presented present VERB bracis-19048 75 12 before before ADP bracis-19048 75 13 the the DET bracis-19048 75 14 description description NOUN bracis-19048 75 15 of of ADP bracis-19048 75 16 the the DET bracis-19048 75 17 clustering cluster VERB bracis-19048 75 18 measures measure NOUN bracis-19048 75 19 . . PUNCT bracis-19048 76 1 let let VERB bracis-19048 76 2 \(x\in \(x\in X bracis-19048 76 3 \mathbb \mathbb PROPN bracis-19048 76 4 { { PUNCT bracis-19048 76 5 r}^{n\times r}^{n\time NOUN bracis-19048 76 6 { { PUNCT bracis-19048 76 7 q}}\ q}}\ NOUN bracis-19048 76 8 ) ) PUNCT bracis-19048 76 9 be be AUX bracis-19048 76 10 a a DET bracis-19048 76 11 matrix matrix NOUN bracis-19048 76 12 of of ADP bracis-19048 76 13 n n DET bracis-19048 76 14 observations observation NOUN bracis-19048 76 15 , , PUNCT bracis-19048 76 16 each each PRON bracis-19048 76 17 represented represent VERB bracis-19048 76 18 by by ADP bracis-19048 76 19 a a DET bracis-19048 76 20 q q ADJ bracis-19048 76 21 - - PUNCT bracis-19048 76 22 dimensional dimensional ADJ bracis-19048 76 23 vector vector NOUN bracis-19048 76 24 \(\mathbf \(\mathbf VERB bracis-19048 76 25 { { PUNCT bracis-19048 76 26 x}_\ell x}_\ell PROPN bracis-19048 76 27 \in \in PROPN bracis-19048 76 28 \mathbb \mathbb PROPN bracis-19048 76 29 { { PUNCT bracis-19048 76 30 r}^q\ r}^q\ NOUN bracis-19048 76 31 ) ) PUNCT bracis-19048 76 32 ; ; PUNCT bracis-19048 76 33 and and CCONJ bracis-19048 76 34 \(u_k(x \(u_k(x NOUN bracis-19048 76 35 ) ) PUNCT bracis-19048 77 1 = = SYM bracis-19048 77 2 \{x_1,x_2,\dotsc \{x_1,x_2,\dotsc PROPN bracis-19048 77 3 , , PUNCT bracis-19048 77 4 x_k\}\ x_k\}\ PROPN bracis-19048 77 5 ) ) PUNCT bracis-19048 77 6 be be AUX bracis-19048 77 7 an an DET bracis-19048 77 8 exhaustive exhaustive ADJ bracis-19048 77 9 partition partition NOUN bracis-19048 77 10 of of ADP bracis-19048 77 11 x x PUNCT bracis-19048 77 12 into into ADP bracis-19048 77 13 k k PROPN bracis-19048 77 14 mutually mutually ADV bracis-19048 77 15 exclusive exclusive ADJ bracis-19048 77 16 clusters cluster NOUN bracis-19048 77 17 \(x_k\ \(x_k\ NUM bracis-19048 77 18 ) ) PUNCT bracis-19048 77 19 with with ADP bracis-19048 77 20 sizes size NOUN bracis-19048 77 21 \(n_k \(n_k ADV bracis-19048 77 22 > > PUNCT bracis-19048 77 23 0,~k=1,\dotsc 0,~k=1,\dotsc NUM bracis-19048 77 24 , , PUNCT bracis-19048 77 25 k\ k\ PROPN bracis-19048 77 26 ) ) PUNCT bracis-19048 77 27 . . PUNCT bracis-19048 78 1 in in ADP bracis-19048 78 2 all all DET bracis-19048 78 3 definitions definition NOUN bracis-19048 78 4 below below ADV bracis-19048 78 5 , , PUNCT bracis-19048 78 6 the the DET bracis-19048 78 7 clustering clustering ADJ bracis-19048 78 8 meta meta NOUN bracis-19048 78 9 - - PUNCT bracis-19048 78 10 features feature NOUN bracis-19048 78 11 are be AUX bracis-19048 78 12 functions function NOUN bracis-19048 78 13 of of ADP bracis-19048 78 14 a a DET bracis-19048 78 15 partition partition NOUN bracis-19048 78 16 \(u_k(x)\ \(u_k(x)\ PROPN bracis-19048 78 17 ) ) PUNCT bracis-19048 78 18 , , PUNCT bracis-19048 78 19 which which PRON bracis-19048 78 20 in in ADP bracis-19048 78 21 the the DET bracis-19048 78 22 case case NOUN bracis-19048 78 23 of of ADP bracis-19048 78 24 this this DET bracis-19048 78 25 paper paper NOUN bracis-19048 78 26 is be AUX bracis-19048 78 27 given give VERB bracis-19048 78 28 by by ADP bracis-19048 78 29 the the DET bracis-19048 78 30 class class NOUN bracis-19048 78 31 labels label NOUN bracis-19048 78 32 . . PUNCT bracis-19048 79 1 the the DET bracis-19048 79 2 \(u_k(x)\ \(u_k(x)\ PROPN bracis-19048 79 3 ) ) PUNCT bracis-19048 79 4 is be AUX bracis-19048 79 5 kept keep VERB bracis-19048 79 6 implicit implicit ADJ bracis-19048 79 7 in in ADP bracis-19048 79 8 the the DET bracis-19048 79 9 definitions definition NOUN bracis-19048 79 10 in in ADP bracis-19048 79 11 order order NOUN bracis-19048 79 12 to to PART bracis-19048 79 13 keep keep VERB bracis-19048 79 14 the the DET bracis-19048 79 15 notation notation NOUN bracis-19048 79 16 cleaner cleaner NOUN bracis-19048 79 17 , , PUNCT bracis-19048 79 18 but but CCONJ bracis-19048 79 19 the the DET bracis-19048 79 20 reader reader NOUN bracis-19048 79 21 should should AUX bracis-19048 79 22 keep keep VERB bracis-19048 79 23 this this DET bracis-19048 79 24 relationship relationship NOUN bracis-19048 79 25 in in ADP bracis-19048 79 26 mind mind NOUN bracis-19048 79 27 . . PUNCT bracis-19048 80 1 let let AUX bracis-19048 80 2 \(\mathbf \(\mathbf VERB bracis-19048 80 3 { { PUNCT bracis-19048 80 4 \bar{x}}_k\ \bar{x}}_k\ PROPN bracis-19048 80 5 ) ) PUNCT bracis-19048 80 6 denote denote VERB bracis-19048 80 7 the the DET bracis-19048 80 8 mean mean ADJ bracis-19048 80 9 point point NOUN bracis-19048 80 10 of of ADP bracis-19048 80 11 all all DET bracis-19048 80 12 observations observation NOUN bracis-19048 80 13 belonging belong VERB bracis-19048 80 14 to to ADP bracis-19048 80 15 cluster cluster NOUN bracis-19048 80 16 \(x_k\ \(x_k\ NOUN bracis-19048 80 17 ) ) PUNCT bracis-19048 80 18 , , PUNCT bracis-19048 80 19 and and CCONJ bracis-19048 80 20 \(\mathbf \(\mathbf VERB bracis-19048 80 21 { { PUNCT bracis-19048 80 22 \bar{\bar{x}}}\ \bar{\bar{x}}}\ PROPN bracis-19048 80 23 ) ) PUNCT bracis-19048 80 24 denote denote VERB bracis-19048 80 25 the the DET bracis-19048 80 26 grand grand ADJ bracis-19048 80 27 mean mean NOUN bracis-19048 80 28 of of ADP bracis-19048 80 29 all all DET bracis-19048 80 30 observations observation NOUN bracis-19048 80 31 in in ADP bracis-19048 80 32 x. x. NOUN bracis-19048 80 33 also also ADV bracis-19048 80 34 , , PUNCT bracis-19048 80 35 let let VERB bracis-19048 80 36 \(dist(\cdot \(dist(\cdot ADV bracis-19048 80 37 , , PUNCT bracis-19048 80 38 \cdot \cdot NOUN bracis-19048 80 39 ) ) PUNCT bracis-19048 80 40 \ \ NOUN bracis-19048 80 41 ) ) PUNCT bracis-19048 80 42 denote denote VERB bracis-19048 80 43 the the DET bracis-19048 80 44 distance distance NOUN bracis-19048 80 45 ( ( PUNCT bracis-19048 80 46 e.g. e.g. ADV bracis-19048 80 47 , , PUNCT bracis-19048 80 48 euclidean euclidean NOUN bracis-19048 80 49 ) ) PUNCT bracis-19048 80 50 between between ADP bracis-19048 80 51 two two NUM bracis-19048 80 52 points point NOUN bracis-19048 80 53 . . PUNCT bracis-19048 81 1 then then ADV bracis-19048 81 2 , , PUNCT bracis-19048 81 3 $ $ SYM bracis-19048 81 4 $ $ SYM bracis-19048 81 5 \begin{aligned \begin{aligne VERB bracis-19048 81 6 } } PUNCT bracis-19048 81 7 \delta \delta ADP bracis-19048 81 8 _ _ PUNCT bracis-19048 81 9 { { PUNCT bracis-19048 81 10 ij ij NOUN bracis-19048 81 11 } } PUNCT bracis-19048 81 12 \triangleq \triangleq PROPN bracis-19048 81 13 \min \min NOUN bracis-19048 81 14 _ _ PROPN bracis-19048 81 15 { { PUNCT bracis-19048 81 16 \begin{array}{c \begin{array}{c PROPN bracis-19048 81 17 } } PUNCT bracis-19048 81 18 \mathbf \mathbf PROPN bracis-19048 81 19 { { PUNCT bracis-19048 81 20 x}\in x}\in PROPN bracis-19048 81 21 x_i\\ x_i\\ PROPN bracis-19048 82 1 \mathbf \mathbf PROPN bracis-19048 82 2 { { PUNCT bracis-19048 82 3 y}\in y}\in PROPN bracis-19048 82 4 x_j x_j SYM bracis-19048 82 5 \end{array \end{array ADJ bracis-19048 82 6 } } PUNCT bracis-19048 82 7 } } PUNCT bracis-19048 82 8 dist(\mathbf dist(\mathbf NOUN bracis-19048 82 9 { { PUNCT bracis-19048 82 10 x},\mathbf x},\mathbf PROPN bracis-19048 82 11 { { PUNCT bracis-19048 82 12 y y PROPN bracis-19048 82 13 } } PUNCT bracis-19048 82 14 ) ) PUNCT bracis-19048 82 15 \end{aligned}$$ \end{aligned}$$ X bracis-19048 82 16 ( ( PUNCT bracis-19048 82 17 1 1 X bracis-19048 82 18 ) ) PUNCT bracis-19048 82 19 denotes denote VERB bracis-19048 82 20 the the DET bracis-19048 82 21 single single ADJ bracis-19048 82 22 - - PUNCT bracis-19048 82 23 linkage linkage NOUN bracis-19048 82 24 distance distance NOUN bracis-19048 82 25 between between ADP bracis-19048 82 26 the the DET bracis-19048 82 27 two two NUM bracis-19048 82 28 clusters cluster NOUN bracis-19048 82 29 \(x_i,~x_j\ \(x_i,~x_j\ PROPN bracis-19048 82 30 ) ) PUNCT bracis-19048 82 31 , , PUNCT bracis-19048 82 32 and and CCONJ bracis-19048 82 33 $ $ SYM bracis-19048 82 34 $ $ SYM bracis-19048 82 35 \begin{aligned \begin{aligne VERB bracis-19048 82 36 } } PUNCT bracis-19048 82 37 \vardelta \vardelta ADP bracis-19048 82 38 _ _ PUNCT bracis-19048 82 39 k k NOUN bracis-19048 82 40 \triangleq \triangleq PROPN bracis-19048 82 41 \max \max PROPN bracis-19048 82 42 _ _ PUNCT bracis-19048 82 43 { { PUNCT bracis-19048 82 44 \mathbf \mathbf PROPN bracis-19048 82 45 { { PUNCT bracis-19048 82 46 x},\mathbf x},\mathbf PROPN bracis-19048 82 47 { { PUNCT bracis-19048 82 48 y}\in y}\in PROPN bracis-19048 82 49 x_k x_k SYM bracis-19048 82 50 } } PUNCT bracis-19048 82 51 dist(\mathbf dist(\mathbf NOUN bracis-19048 82 52 { { PUNCT bracis-19048 82 53 x},\mathbf x},\mathbf PROPN bracis-19048 82 54 { { PUNCT bracis-19048 82 55 y y PROPN bracis-19048 82 56 } } PUNCT bracis-19048 82 57 ) ) PUNCT bracis-19048 82 58 \end{aligned}$$ \end{aligned}$$ PROPN bracis-19048 82 59 ( ( PUNCT bracis-19048 82 60 2 2 NUM bracis-19048 82 61 ) ) PUNCT bracis-19048 82 62 represents represent VERB bracis-19048 82 63 the the DET bracis-19048 82 64 diameter diameter NOUN bracis-19048 82 65 of of ADP bracis-19048 82 66 a a DET bracis-19048 82 67 cluster cluster NOUN bracis-19048 82 68 \(x_k\ \(x_k\ NOUN bracis-19048 82 69 ) ) PUNCT bracis-19048 82 70 . . PUNCT bracis-19048 83 1 additionally additionally ADV bracis-19048 83 2 , , PUNCT bracis-19048 83 3 consider consider VERB bracis-19048 83 4 some some DET bracis-19048 83 5 vectors vector NOUN bracis-19048 83 6 of of ADP bracis-19048 83 7 distances distance NOUN bracis-19048 83 8 . . PUNCT bracis-19048 84 1 let let AUX bracis-19048 84 2 \(\mathbf \(\mathbf VERB bracis-19048 84 3 { { PUNCT bracis-19048 84 4 d}^+\in d}^+\in VERB bracis-19048 84 5 \mathbb \mathbb PROPN bracis-19048 84 6 { { PUNCT bracis-19048 84 7 r}^{n_w}\ r}^{n_w}\ NOUN bracis-19048 84 8 ) ) PUNCT bracis-19048 84 9 be be AUX bracis-19048 84 10 a a DET bracis-19048 84 11 vector vector NOUN bracis-19048 84 12 of of ADP bracis-19048 84 13 all all DET bracis-19048 84 14 \(n_w\ \(n_w\ NOUN bracis-19048 84 15 ) ) PUNCT bracis-19048 84 16 within within ADP bracis-19048 84 17 - - PUNCT bracis-19048 84 18 cluster cluster NOUN bracis-19048 84 19 distances distance NOUN bracis-19048 84 20 ( ( PUNCT bracis-19048 84 21 i.e. i.e. X bracis-19048 84 22 , , PUNCT bracis-19048 84 23 all all DET bracis-19048 84 24 distances distance NOUN bracis-19048 84 25 between between ADP bracis-19048 84 26 pairs pair NOUN bracis-19048 84 27 of of ADP bracis-19048 84 28 points point NOUN bracis-19048 84 29 having have VERB bracis-19048 84 30 equal equal ADJ bracis-19048 84 31 class class NOUN bracis-19048 84 32 labels label NOUN bracis-19048 84 33 in in ADP bracis-19048 84 34 the the DET bracis-19048 84 35 data datum NOUN bracis-19048 84 36 ) ) PUNCT bracis-19048 84 37 , , PUNCT bracis-19048 84 38 \(\mathbf \(\mathbf VERB bracis-19048 84 39 { { PUNCT bracis-19048 84 40 d}^-\in d}^-\in NOUN bracis-19048 84 41 \mathbb \mathbb PROPN bracis-19048 84 42 { { PUNCT bracis-19048 84 43 r}^{n_b}\ r}^{n_b}\ NOUN bracis-19048 84 44 ) ) PUNCT bracis-19048 84 45 be be VERB bracis-19048 84 46 a a DET bracis-19048 84 47 vector vector NOUN bracis-19048 84 48 containing contain VERB bracis-19048 84 49 all all DET bracis-19048 84 50 \(n_b\ \(n_b\ NOUN bracis-19048 84 51 ) ) PUNCT bracis-19048 84 52 between between ADP bracis-19048 84 53 - - PUNCT bracis-19048 84 54 cluster cluster NOUN bracis-19048 84 55 distances distance NOUN bracis-19048 84 56 ( ( PUNCT bracis-19048 84 57 between between ADP bracis-19048 84 58 pairs pair NOUN bracis-19048 84 59 of of ADP bracis-19048 84 60 points point NOUN bracis-19048 84 61 having have VERB bracis-19048 84 62 different different ADJ bracis-19048 84 63 class class NOUN bracis-19048 84 64 labels label NOUN bracis-19048 84 65 ) ) PUNCT bracis-19048 84 66 , , PUNCT bracis-19048 84 67 and and CCONJ bracis-19048 84 68 \(\mathbf \(\mathbf VERB bracis-19048 84 69 { { PUNCT bracis-19048 84 70 d}^\bullet d}^\bullet PROPN bracis-19048 84 71 \in \in PROPN bracis-19048 84 72 \mathbb \mathbb PROPN bracis-19048 84 73 { { PUNCT bracis-19048 84 74 r}^{n_t}\ r}^{n_t}\ NOUN bracis-19048 84 75 ) ) PUNCT bracis-19048 84 76 denote denote VERB bracis-19048 84 77 a a DET bracis-19048 84 78 vector vector NOUN bracis-19048 84 79 containing contain VERB bracis-19048 84 80 the the DET bracis-19048 84 81 distances distance NOUN bracis-19048 84 82 between between ADP bracis-19048 84 83 all all DET bracis-19048 84 84 \(n_t\ \(n_t\ NOUN bracis-19048 84 85 ) ) PUNCT bracis-19048 84 86 pairs pair NOUN bracis-19048 84 87 of of ADP bracis-19048 84 88 points point NOUN bracis-19048 84 89 in in ADP bracis-19048 84 90 x x NOUN bracis-19048 84 91 , , PUNCT bracis-19048 84 92 with with ADP bracis-19048 84 93 \(n_t \(n_t NOUN bracis-19048 84 94 = = SYM bracis-19048 84 95 n_w n_w PUNCT bracis-19048 84 96 + + NUM bracis-19048 84 97 n_b\ n_b\ NOUN bracis-19048 84 98 ) ) PUNCT bracis-19048 84 99 . . PUNCT bracis-19048 85 1 finally finally ADV bracis-19048 85 2 , , PUNCT bracis-19048 85 3 let let VERB bracis-19048 85 4 the the DET bracis-19048 85 5 following follow VERB bracis-19048 85 6 quantities quantity NOUN bracis-19048 85 7 be be AUX bracis-19048 85 8 defined define VERB bracis-19048 85 9 : : PUNCT bracis-19048 85 10 the the DET bracis-19048 85 11 within within ADP bracis-19048 85 12 - - PUNCT bracis-19048 85 13 groups group NOUN bracis-19048 85 14 sum sum NOUN bracis-19048 85 15 of of ADP bracis-19048 85 16 squares square NOUN bracis-19048 85 17 , , PUNCT bracis-19048 86 1 $ $ SYM bracis-19048 86 2 $ $ SYM bracis-19048 86 3 \begin{aligned \begin{aligne VERB bracis-19048 86 4 } } PUNCT bracis-19048 86 5 wgss wgss NOUN bracis-19048 86 6 = = SYM bracis-19048 86 7 \sum \sum NOUN bracis-19048 86 8 _ _ PUNCT bracis-19048 86 9 { { PUNCT bracis-19048 86 10 k=1}^k\sum k=1}^k\sum ADP bracis-19048 86 11 _ _ PRON bracis-19048 86 12 { { PUNCT bracis-19048 86 13 \mathbf \mathbf NOUN bracis-19048 86 14 { { PUNCT bracis-19048 86 15 x}_i\in x}_i\in ADJ bracis-19048 86 16 x_k x_k SYM bracis-19048 86 17 } } PUNCT bracis-19048 86 18 \left \left PROPN bracis-19048 86 19 [ [ PUNCT bracis-19048 86 20 dist(\mathbf dist(\mathbf NOUN bracis-19048 86 21 { { PUNCT bracis-19048 86 22 x}_i x}_i PROPN bracis-19048 86 23 , , PUNCT bracis-19048 86 24 \mathbf \mathbf PROPN bracis-19048 86 25 { { PUNCT bracis-19048 86 26 \bar{x}}_k)\right \bar{x}}_k)\right NOUN bracis-19048 86 27 ] ] X bracis-19048 86 28 ^2 ^2 NUM bracis-19048 86 29 , , PUNCT bracis-19048 86 30 \end{aligned}$$ \end{aligned}$$ X bracis-19048 86 31 ( ( PUNCT bracis-19048 86 32 3 3 NUM bracis-19048 86 33 ) ) PUNCT bracis-19048 86 34 and and CCONJ bracis-19048 86 35 the the DET bracis-19048 86 36 between between NOUN bracis-19048 86 37 - - PUNCT bracis-19048 86 38 groups group NOUN bracis-19048 86 39 sum sum NOUN bracis-19048 86 40 of of ADP bracis-19048 86 41 squares square NOUN bracis-19048 86 42 , , PUNCT bracis-19048 87 1 $ $ SYM bracis-19048 87 2 $ $ SYM bracis-19048 87 3 \begin{aligned \begin{aligne VERB bracis-19048 87 4 } } PUNCT bracis-19048 87 5 bgss bgss NOUN bracis-19048 87 6 = = PUNCT bracis-19048 87 7 \sum \sum NOUN bracis-19048 87 8 _ _ PUNCT bracis-19048 87 9 { { PUNCT bracis-19048 87 10 k=1}^k k=1}^k PROPN bracis-19048 87 11 n_k\left n_k\left PROPN bracis-19048 87 12 [ [ PUNCT bracis-19048 87 13 dist(\mathbf dist(\mathbf PROPN bracis-19048 87 14 { { PUNCT bracis-19048 87 15 \bar{x}}_k \bar{x}}_k PROPN bracis-19048 87 16 , , PUNCT bracis-19048 87 17 \mathbf \mathbf PROPN bracis-19048 87 18 { { PUNCT bracis-19048 87 19 \bar{\bar{x}}})\right \bar{\bar{x}}})\right PROPN bracis-19048 87 20 ] ] PUNCT bracis-19048 87 21 ^2 ^2 NUM bracis-19048 87 22 . . PUNCT bracis-19048 88 1 \end{aligned}$$ \end{aligned}$$ X bracis-19048 88 2 ( ( PUNCT bracis-19048 88 3 4 4 NUM bracis-19048 88 4 ) ) PUNCT bracis-19048 88 5 given give VERB bracis-19048 88 6 the the DET bracis-19048 88 7 preceding precede VERB bracis-19048 88 8 definitions definition NOUN bracis-19048 88 9 , , PUNCT bracis-19048 88 10 the the DET bracis-19048 88 11 clustering clustering ADJ bracis-19048 88 12 meta meta NOUN bracis-19048 88 13 - - PUNCT bracis-19048 88 14 features feature NOUN bracis-19048 88 15 used use VERB bracis-19048 88 16 in in ADP bracis-19048 88 17 this this DET bracis-19048 88 18 work work NOUN bracis-19048 88 19 are be AUX bracis-19048 88 20 formalized formalize VERB bracis-19048 88 21 as as SCONJ bracis-19048 88 22 follows follow VERB bracis-19048 88 23 : : PUNCT bracis-19048 88 24 dunn dunn PROPN bracis-19048 88 25 ’s ’s PART bracis-19048 88 26 separation separation NOUN bracis-19048 88 27 index index NOUN bracis-19048 88 28 ( ( PUNCT bracis-19048 88 29 vdu vdu PROPN bracis-19048 88 30 ) ) PUNCT bracis-19048 89 1 [ [ X bracis-19048 89 2 14 14 NUM bracis-19048 89 3 ] ] SYM bracis-19048 89 4 : : PUNCT bracis-19048 89 5 $ $ SYM bracis-19048 89 6 $ $ SYM bracis-19048 89 7 \begin{aligned \begin{aligne VERB bracis-19048 89 8 } } PUNCT bracis-19048 89 9 vdu vdu PROPN bracis-19048 89 10 = = SYM bracis-19048 89 11 \min \min PROPN bracis-19048 89 12 _ _ PUNCT bracis-19048 89 13 { { PUNCT bracis-19048 89 14 \begin{array}{c \begin{array}{c PROPN bracis-19048 89 15 } } PUNCT bracis-19048 89 16 i i PROPN bracis-19048 89 17 , , PUNCT bracis-19048 89 18 j\in j\in PROPN bracis-19048 90 1 [ [ X bracis-19048 90 2 1,k]\\ 1,k]\\ NUM bracis-19048 90 3 i\ne i\ne PROPN bracis-19048 90 4 j j PROPN bracis-19048 91 1 \end{array}}\frac{\delta \end{array}}\frac{\delta ADP bracis-19048 91 2 _ _ PROPN bracis-19048 91 3 { { PUNCT bracis-19048 91 4 ij}}{\max ij}}{\max ADV bracis-19048 91 5 \limits \limit NOUN bracis-19048 91 6 _ _ PRON bracis-19048 91 7 { { PUNCT bracis-19048 91 8 k\in k\in VERB bracis-19048 91 9 [ [ X bracis-19048 91 10 1,k]}\vardelta 1,k]}\vardelta NUM bracis-19048 91 11 _ _ PUNCT bracis-19048 91 12 k k X bracis-19048 91 13 } } PUNCT bracis-19048 91 14 . . PUNCT bracis-19048 92 1 \end{aligned}$$ \end{aligned}$$ PROPN bracis-19048 92 2 ( ( PUNCT bracis-19048 92 3 5 5 NUM bracis-19048 92 4 ) ) PUNCT bracis-19048 92 5 davies davy NOUN bracis-19048 92 6 - - PUNCT bracis-19048 92 7 bouldin bouldin NOUN bracis-19048 92 8 index index NOUN bracis-19048 92 9 ( ( PUNCT bracis-19048 92 10 vdb vdb NOUN bracis-19048 92 11 ) ) PUNCT bracis-19048 93 1 [ [ X bracis-19048 93 2 11 11 NUM bracis-19048 93 3 ] ] SYM bracis-19048 93 4 : : PUNCT bracis-19048 93 5 $ $ SYM bracis-19048 93 6 $ $ SYM bracis-19048 93 7 \begin{aligned \begin{aligne VERB bracis-19048 93 8 } } PUNCT bracis-19048 93 9 vdb vdb NOUN bracis-19048 93 10 = = SYM bracis-19048 93 11 \frac{1}{k}\sum \frac{1}{k}\sum NOUN bracis-19048 93 12 _ _ PRON bracis-19048 93 13 { { PUNCT bracis-19048 93 14 i=1}^k\max i=1}^k\max ADJ bracis-19048 93 15 _ _ PUNCT bracis-19048 93 16 { { PUNCT bracis-19048 93 17 j\ne j\ne PROPN bracis-19048 93 18 i}\frac{\vardelta i}\frac{\vardelta PROPN bracis-19048 94 1 _ _ PUNCT bracis-19048 95 1 i i PRON bracis-19048 95 2 + + PUNCT bracis-19048 95 3 \vardelta \vardelta X bracis-19048 95 4 _ _ NOUN bracis-19048 95 5 j}{dist(\mathbf j}{dist(\mathbf NOUN bracis-19048 95 6 { { PUNCT bracis-19048 95 7 \bar{x}}_i,\mathbf \bar{x}}_i,\mathbf NOUN bracis-19048 95 8 { { PUNCT bracis-19048 95 9 \bar{x}}_j \bar{x}}_j NOUN bracis-19048 95 10 ) ) PUNCT bracis-19048 95 11 } } PUNCT bracis-19048 95 12 . . PUNCT bracis-19048 96 1 \end{aligned}$$ \end{aligned}$$ PROPN bracis-19048 96 2 ( ( PUNCT bracis-19048 96 3 6 6 NUM bracis-19048 96 4 ) ) PUNCT bracis-19048 96 5 baker baker PROPN bracis-19048 96 6 - - PUNCT bracis-19048 96 7 hubert hubert PROPN bracis-19048 96 8 index index NOUN bracis-19048 96 9 ( ( PUNCT bracis-19048 96 10 \(\vargamma \(\vargamma PROPN bracis-19048 96 11 \ \ PROPN bracis-19048 96 12 ) ) PUNCT bracis-19048 96 13 ) ) PUNCT bracis-19048 97 1 [ [ X bracis-19048 97 2 1 1 NUM bracis-19048 97 3 ] ] PUNCT bracis-19048 97 4 : : PUNCT bracis-19048 97 5 let let VERB bracis-19048 97 6 $ $ SYM bracis-19048 97 7 $ $ SYM bracis-19048 97 8 \begin{aligned \begin{aligne VERB bracis-19048 97 9 } } PUNCT bracis-19048 97 10 s^+ s^+ ADJ bracis-19048 97 11 = = PUNCT bracis-19048 97 12 \sum \sum NOUN bracis-19048 97 13 _ _ PUNCT bracis-19048 97 14 { { PUNCT bracis-19048 97 15 \forall \forall DET bracis-19048 97 16 d_i\in d_i\in NOUN bracis-19048 97 17 \mathbf \mathbf PROPN bracis-19048 97 18 { { PUNCT bracis-19048 97 19 d}^+}\sum d}^+}\sum VERB bracis-19048 97 20 _ _ PRON bracis-19048 97 21 { { PUNCT bracis-19048 97 22 \forall \forall PRON bracis-19048 97 23 d_j\in d_j\in VERB bracis-19048 97 24 \mathbf \mathbf PROPN bracis-19048 97 25 { { PUNCT bracis-19048 97 26 d}^-}one(d_i d}^-}one(d_i NOUN bracis-19048 97 27 < < NOUN bracis-19048 97 28 d_j d_j NOUN bracis-19048 97 29 ) ) PUNCT bracis-19048 97 30 \quad \quad PROPN bracis-19048 97 31 \mathrm \mathrm PROPN bracis-19048 97 32 { { PUNCT bracis-19048 97 33 and}\quad and}\quad PROPN bracis-19048 97 34 s^= s^= ADJ bracis-19048 97 35 \sum \sum NOUN bracis-19048 97 36 _ _ PUNCT bracis-19048 97 37 { { PUNCT bracis-19048 97 38 \forall \forall DET bracis-19048 97 39 d_i\in d_i\in NOUN bracis-19048 97 40 \mathbf \mathbf PROPN bracis-19048 97 41 { { PUNCT bracis-19048 97 42 d}^+}\sum d}^+}\sum VERB bracis-19048 97 43 _ _ PRON bracis-19048 97 44 { { PUNCT bracis-19048 97 45 \forall \forall PRON bracis-19048 97 46 d_j\in d_j\in VERB bracis-19048 97 47 \mathbf \mathbf PROPN bracis-19048 97 48 { { PUNCT bracis-19048 97 49 d}^-}one(d_i d}^-}one(d_i NOUN bracis-19048 97 50 > > PUNCT bracis-19048 97 51 d_j d_j NOUN bracis-19048 97 52 ) ) PUNCT bracis-19048 97 53 \end{aligned}$$ \end{aligned}$$ X bracis-19048 97 54 ( ( PUNCT bracis-19048 97 55 7 7 NUM bracis-19048 97 56 ) ) PUNCT bracis-19048 97 57 where where SCONJ bracis-19048 97 58 one(condition one(condition NOUN bracis-19048 97 59 ) ) PUNCT bracis-19048 97 60 is be AUX bracis-19048 97 61 a a DET bracis-19048 97 62 function function NOUN bracis-19048 97 63 that that PRON bracis-19048 97 64 returns return VERB bracis-19048 97 65 1 1 NUM bracis-19048 97 66 if if SCONJ bracis-19048 97 67 the the DET bracis-19048 97 68 condition condition NOUN bracis-19048 97 69 is be AUX bracis-19048 97 70 true true ADJ bracis-19048 97 71 and and CCONJ bracis-19048 97 72 0 0 NUM bracis-19048 97 73 otherwise otherwise ADV bracis-19048 97 74 ; ; PUNCT bracis-19048 97 75 then then ADV bracis-19048 97 76 : : PUNCT bracis-19048 97 77 $ $ SYM bracis-19048 97 78 $ $ SYM bracis-19048 97 79 \begin{aligned \begin{aligne VERB bracis-19048 97 80 } } PUNCT bracis-19048 97 81 \vargamma \vargamma NOUN bracis-19048 97 82 = = NOUN bracis-19048 97 83 \frac{s^+ \frac{s^+ NOUN bracis-19048 97 84 s^-}{s^+ s^-}{s^+ VERB bracis-19048 97 85 + + CCONJ bracis-19048 97 86 s^- s^- ADJ bracis-19048 97 87 } } PUNCT bracis-19048 97 88 . . PUNCT bracis-19048 98 1 \end{aligned}$$ \end{aligned}$$ PROPN bracis-19048 98 2 ( ( PUNCT bracis-19048 98 3 8) 8) NUM bracis-19048 98 4 tie tie NOUN bracis-19048 98 5 - - PUNCT bracis-19048 98 6 corrected correct VERB bracis-19048 98 7 kendall kendall PROPN bracis-19048 98 8 tau tau PROPN bracis-19048 98 9 ( ( PUNCT bracis-19048 98 10 \(\tau \(\tau ADV bracis-19048 98 11 \ \ NUM bracis-19048 98 12 ) ) PUNCT bracis-19048 98 13 ) ) PUNCT bracis-19048 99 1 [ [ X bracis-19048 99 2 12 12 NUM bracis-19048 99 3 ] ] X bracis-19048 99 4 : : PUNCT bracis-19048 99 5 $ $ SYM bracis-19048 99 6 $ $ SYM bracis-19048 99 7 \begin{aligned \begin{aligne VERB bracis-19048 99 8 } } PUNCT bracis-19048 99 9 \tau \tau PUNCT bracis-19048 99 10 = = PUNCT bracis-19048 99 11 \frac{s^+ \frac{s^+ NUM bracis-19048 99 12 s^-}{\sqrt{n_wn_bn_t(n_t-1)/2 s^-}{\sqrt{n_wn_bn_t(n_t-1)/2 PROPN bracis-19048 99 13 } } PUNCT bracis-19048 99 14 } } PUNCT bracis-19048 99 15 . . PUNCT bracis-19048 100 1 \end{aligned}$$ \end{aligned}$$ PROPN bracis-19048 100 2 ( ( PUNCT bracis-19048 100 3 9 9 NUM bracis-19048 100 4 ) ) PUNCT bracis-19048 100 5 ray ray NOUN bracis-19048 100 6 - - PUNCT bracis-19048 100 7 turi turi NOUN bracis-19048 100 8 index index NOUN bracis-19048 100 9 ( ( PUNCT bracis-19048 100 10 \(\nu \(\nu PROPN bracis-19048 100 11 \ \ PROPN bracis-19048 100 12 ) ) PUNCT bracis-19048 100 13 ) ) PUNCT bracis-19048 101 1 [ [ X bracis-19048 101 2 28 28 NUM bracis-19048 101 3 ] ] X bracis-19048 101 4 : : PUNCT bracis-19048 101 5 $ $ SYM bracis-19048 101 6 $ $ SYM bracis-19048 101 7 \begin{aligned \begin{aligne VERB bracis-19048 101 8 } } PUNCT bracis-19048 101 9 \nu \nu PROPN bracis-19048 101 10 = = SYM bracis-19048 101 11 \frac{1}{n}\frac{wgss}{\min \frac{1}{n}\frac{wgss}{\min PROPN bracis-19048 101 12 \limits \limit NOUN bracis-19048 101 13 _ _ PRON bracis-19048 101 14 { { PUNCT bracis-19048 101 15 i\ne i\ne PROPN bracis-19048 101 16 j j PROPN bracis-19048 101 17 } } PUNCT bracis-19048 101 18 \delta \delta VERB bracis-19048 101 19 _ _ PRON bracis-19048 101 20 { { PUNCT bracis-19048 101 21 ij}^2 ij}^2 NOUN bracis-19048 101 22 } } PUNCT bracis-19048 101 23 . . PUNCT bracis-19048 102 1 \end{aligned}$$ \end{aligned}$$ PROPN bracis-19048 102 2 ( ( PUNCT bracis-19048 102 3 10 10 NUM bracis-19048 102 4 ) ) PUNCT bracis-19048 102 5 mean mean VERB bracis-19048 102 6 inter inter ADJ bracis-19048 102 7 - - NOUN bracis-19048 102 8 centroid centroid ADJ bracis-19048 102 9 distance distance NOUN bracis-19048 102 10 ( ( PUNCT bracis-19048 102 11 int int NOUN bracis-19048 102 12 ) ) PUNCT bracis-19048 102 13     SPACE bracis-19048 103 1 [ [ X bracis-19048 103 2 2 2 NUM bracis-19048 103 3 ] ] NOUN bracis-19048 103 4 : : PUNCT bracis-19048 103 5 $ $ SYM bracis-19048 103 6 $ $ SYM bracis-19048 103 7 \begin{aligned \begin{aligne VERB bracis-19048 103 8 } } PUNCT bracis-19048 103 9 int int NOUN bracis-19048 103 10 = = NOUN bracis-19048 103 11 \frac{2}{k(k-1)}\sum \frac{2}{k(k-1)}\sum NOUN bracis-19048 103 12 _ _ PUNCT bracis-19048 103 13 { { PUNCT bracis-19048 103 14 k=1}^{k-1 k=1}^{k-1 NOUN bracis-19048 103 15 } } PUNCT bracis-19048 103 16 dist(\mathbf dist(\mathbf NOUN bracis-19048 103 17 { { PUNCT bracis-19048 103 18 \bar{x}}_k \bar{x}}_k PROPN bracis-19048 103 19 , , PUNCT bracis-19048 103 20 \mathbf \mathbf PROPN bracis-19048 103 21 { { PUNCT bracis-19048 103 22 \bar{x}}_{k+1 \bar{x}}_{k+1 ADJ bracis-19048 103 23 } } PUNCT bracis-19048 103 24 ) ) PUNCT bracis-19048 103 25 . . PUNCT bracis-19048 104 1 \end{aligned}$$ \end{aligned}$$ PROPN bracis-19048 104 2 ( ( PUNCT bracis-19048 104 3 11 11 NUM bracis-19048 104 4 ) ) PUNCT bracis-19048 104 5 global global ADJ bracis-19048 104 6 silhouette silhouette NOUN bracis-19048 104 7 index index NOUN bracis-19048 104 8 ( ( PUNCT bracis-19048 104 9 sil sil PROPN bracis-19048 104 10 ) ) PUNCT bracis-19048 104 11     SPACE bracis-19048 105 1 [ [ X bracis-19048 105 2 32 32 NUM bracis-19048 105 3 ] ] PUNCT bracis-19048 105 4 : : PUNCT bracis-19048 105 5 for for ADP bracis-19048 105 6 a a DET bracis-19048 105 7 given give VERB bracis-19048 105 8 point point NOUN bracis-19048 105 9 \(\mathbf \(\mathbf X bracis-19048 105 10 { { PUNCT bracis-19048 105 11 x}\in x}\in PROPN bracis-19048 105 12 { { PUNCT bracis-19048 105 13 x_k}\ x_k}\ NOUN bracis-19048 105 14 ) ) PUNCT bracis-19048 105 15 , , PUNCT bracis-19048 105 16 let let VERB bracis-19048 105 17 \(a(\mathbf \(a(\mathbf VERB bracis-19048 105 18 { { PUNCT bracis-19048 105 19 x})\ x})\ NOUN bracis-19048 105 20 ) ) PUNCT bracis-19048 105 21 denote denote VERB bracis-19048 105 22 the the DET bracis-19048 105 23 mean mean ADJ bracis-19048 105 24 distance distance NOUN bracis-19048 105 25 between between ADP bracis-19048 105 26 this this DET bracis-19048 105 27 point point NOUN bracis-19048 105 28 and and CCONJ bracis-19048 105 29 all all DET bracis-19048 105 30 other other ADJ bracis-19048 105 31 points point NOUN bracis-19048 105 32 belonging belong VERB bracis-19048 105 33 to to ADP bracis-19048 105 34 the the DET bracis-19048 105 35 same same ADJ bracis-19048 105 36 cluster cluster NOUN bracis-19048 105 37 , , PUNCT bracis-19048 105 38 \(x_k\ \(x_k\ PROPN bracis-19048 105 39 ) ) PUNCT bracis-19048 105 40 ; ; PUNCT bracis-19048 105 41 \(\mathfrak \(\mathfrak PROPN bracis-19048 105 42 { { PUNCT bracis-19048 105 43 d}(\mathbf d}(\mathbf NOUN bracis-19048 105 44 { { PUNCT bracis-19048 105 45 x},k^\prime x},k^\prime PROPN bracis-19048 105 46 ) ) PUNCT bracis-19048 105 47 \ \ NOUN bracis-19048 105 48 ) ) PUNCT bracis-19048 105 49 denote denote VERB bracis-19048 105 50 the the DET bracis-19048 105 51 mean mean ADJ bracis-19048 105 52 distance distance NOUN bracis-19048 105 53 between between ADP bracis-19048 105 54 this this DET bracis-19048 105 55 point point NOUN bracis-19048 105 56 and and CCONJ bracis-19048 105 57 all all DET bracis-19048 105 58 points point NOUN bracis-19048 105 59 belonging belong VERB bracis-19048 105 60 to to ADP bracis-19048 105 61 a a DET bracis-19048 105 62 distinct distinct ADJ bracis-19048 105 63 cluster cluster NOUN bracis-19048 105 64 \(x_{k^\prime \(x_{k^\prime NOUN bracis-19048 105 65 } } PUNCT bracis-19048 105 66 \ \ NOUN bracis-19048 105 67 ) ) PUNCT bracis-19048 106 1 ( ( PUNCT bracis-19048 106 2 \(k^\prime \(k^\prime PROPN bracis-19048 106 3 \ne \ne PROPN bracis-19048 106 4 k\ k\ PROPN bracis-19048 106 5 ) ) PUNCT bracis-19048 106 6 ) ) PUNCT bracis-19048 106 7 ; ; PUNCT bracis-19048 106 8 and and CCONJ bracis-19048 106 9 $ $ SYM bracis-19048 106 10 $ $ SYM bracis-19048 106 11 \begin{aligned \begin{aligne VERB bracis-19048 106 12 } } PUNCT bracis-19048 106 13 b(\mathbf b(\mathbf NOUN bracis-19048 106 14 { { PUNCT bracis-19048 106 15 x x NOUN bracis-19048 106 16 } } PUNCT bracis-19048 106 17 ) ) PUNCT bracis-19048 107 1 = = VERB bracis-19048 107 2 \min \min NOUN bracis-19048 107 3 _ _ PUNCT bracis-19048 107 4 { { PUNCT bracis-19048 107 5 k^\prime k^\prime PROPN bracis-19048 107 6 \ne \ne PROPN bracis-19048 107 7 k k PROPN bracis-19048 107 8 } } PUNCT bracis-19048 107 9 \mathfrak \mathfrak PROPN bracis-19048 107 10 { { PUNCT bracis-19048 107 11 d}(\mathbf d}(\mathbf NOUN bracis-19048 107 12 { { PUNCT bracis-19048 107 13 x},k^\prime x},k^\prime PROPN bracis-19048 107 14 ) ) PUNCT bracis-19048 107 15 . . PUNCT bracis-19048 108 1 \end{aligned}$$ \end{aligned}$$ PROPN bracis-19048 108 2 ( ( PUNCT bracis-19048 108 3 12 12 NUM bracis-19048 108 4 ) ) PUNCT bracis-19048 108 5 the the DET bracis-19048 108 6 silhouette silhouette NOUN bracis-19048 108 7 width width NOUN bracis-19048 108 8 of of ADP bracis-19048 108 9 point point NOUN bracis-19048 108 10 \(\mathbf \(\mathbf X bracis-19048 108 11 { { PUNCT bracis-19048 108 12 x}\ x}\ PROPN bracis-19048 108 13 ) ) PUNCT bracis-19048 108 14 is be AUX bracis-19048 108 15 then then ADV bracis-19048 108 16 calculated calculate VERB bracis-19048 108 17 as as ADP bracis-19048 108 18 $ $ SYM bracis-19048 108 19 $ $ SYM bracis-19048 108 20 \begin{aligned \begin{aligne VERB bracis-19048 108 21 } } PUNCT bracis-19048 108 22 \mathcal \mathcal ADJ bracis-19048 108 23 { { PUNCT bracis-19048 108 24 s}(\mathbf s}(\mathbf NOUN bracis-19048 108 25 { { PUNCT bracis-19048 108 26 x x NOUN bracis-19048 108 27 } } PUNCT bracis-19048 108 28 ) ) PUNCT bracis-19048 109 1 = = VERB bracis-19048 109 2 \frac{b(\mathbf \frac{b(\mathbf VERB bracis-19048 109 3 { { PUNCT bracis-19048 109 4 x x NOUN bracis-19048 109 5 } } PUNCT bracis-19048 109 6 ) ) PUNCT bracis-19048 109 7 a(\mathbf a(\mathbf PROPN bracis-19048 109 8 { { PUNCT bracis-19048 109 9 x})}{\max x})}{\max PROPN bracis-19048 109 10 ( ( PUNCT bracis-19048 109 11 b(\mathbf b(\mathbf NOUN bracis-19048 109 12 { { PUNCT bracis-19048 109 13 x x NOUN bracis-19048 109 14 } } PUNCT bracis-19048 109 15 ) ) PUNCT bracis-19048 109 16 , , PUNCT bracis-19048 109 17 a(\mathbf a(\mathbf PROPN bracis-19048 109 18 { { PUNCT bracis-19048 109 19 x x NOUN bracis-19048 109 20 } } PUNCT bracis-19048 109 21 ) ) PUNCT bracis-19048 109 22 ) ) PUNCT bracis-19048 109 23 } } PUNCT bracis-19048 109 24 , , PUNCT bracis-19048 109 25 \end{aligned}$$ \end{aligned}$$ X bracis-19048 109 26 ( ( PUNCT bracis-19048 109 27 13 13 NUM bracis-19048 109 28 ) ) PUNCT bracis-19048 109 29 and and CCONJ bracis-19048 109 30 the the DET bracis-19048 109 31 global global ADJ bracis-19048 109 32 silhouette silhouette NOUN bracis-19048 109 33 index index NOUN bracis-19048 109 34 can can AUX bracis-19048 109 35 be be AUX bracis-19048 109 36 calculated calculate VERB bracis-19048 109 37 as as ADP bracis-19048 109 38 $ $ SYM bracis-19048 109 39 $ $ SYM bracis-19048 109 40 \begin{aligned \begin{aligne VERB bracis-19048 109 41 } } PUNCT bracis-19048 109 42 sil sil NOUN bracis-19048 109 43 = = PUNCT bracis-19048 109 44 \frac{1}{k}\sum \frac{1}{k}\sum NOUN bracis-19048 109 45 _ _ PRON bracis-19048 109 46 { { PUNCT bracis-19048 109 47 k=1}^k\sum k=1}^k\sum ADP bracis-19048 109 48 _ _ PRON bracis-19048 109 49 { { PUNCT bracis-19048 109 50 \forall \forall DET bracis-19048 109 51 \mathbf \mathbf PROPN bracis-19048 109 52 { { PUNCT bracis-19048 109 53 x}\in x}\in PROPN bracis-19048 109 54 x_k}\frac{\mathcal x_k}\frac{\mathcal PROPN bracis-19048 109 55 { { PUNCT bracis-19048 109 56 s}(\mathbf s}(\mathbf NOUN bracis-19048 109 57 { { PUNCT bracis-19048 109 58 x})}{n_k x})}{n_k INTJ bracis-19048 109 59 } } PUNCT bracis-19048 109 60 . . PUNCT bracis-19048 110 1 \end{aligned}$$ \end{aligned}$$ PROPN bracis-19048 110 2 ( ( PUNCT bracis-19048 110 3 14 14 NUM bracis-19048 110 4 ) ) PUNCT bracis-19048 110 5 point point NOUN bracis-19048 110 6 biserial biserial ADJ bracis-19048 110 7 index index NOUN bracis-19048 110 8 ( ( PUNCT bracis-19048 110 9 pb pb NOUN bracis-19048 110 10 ) ) PUNCT bracis-19048 111 1 [ [ X bracis-19048 111 2 12 12 NUM bracis-19048 111 3 ] ] X bracis-19048 111 4 : : PUNCT bracis-19048 111 5 $ $ SYM bracis-19048 111 6 $ $ SYM bracis-19048 111 7 \begin{aligned \begin{aligne VERB bracis-19048 111 8 } } PUNCT bracis-19048 111 9 pb pb ADP bracis-19048 111 10 = = SYM bracis-19048 111 11 \left \left PROPN bracis-19048 111 12 ( ( PUNCT bracis-19048 111 13 \frac{|\mathbf \frac{|\mathbf PUNCT bracis-19048 111 14 { { PUNCT bracis-19048 111 15 d}^+|_1}{n_w d}^+|_1}{n_w NOUN bracis-19048 111 16 } } PUNCT bracis-19048 111 17 \frac{|\mathbf \frac{|\mathbf NOUN bracis-19048 111 18 { { PUNCT bracis-19048 111 19 d}^-|_1}{n_b}\right d}^-|_1}{n_b}\right PROPN bracis-19048 111 20 ) ) PUNCT bracis-19048 111 21 \sqrt{\frac{n_wn_b}{n_t^2 \sqrt{\frac{n_wn_b}{n_t^2 PROPN bracis-19048 111 22 } } PUNCT bracis-19048 111 23 } } PUNCT bracis-19048 111 24 . . PUNCT bracis-19048 111 25 \end{aligned}$$ \end{aligned}$$ PROPN bracis-19048 111 26 ( ( PUNCT bracis-19048 111 27 15 15 NUM bracis-19048 111 28 ) ) PUNCT bracis-19048 111 29 with with ADP bracis-19048 111 30 \(|\cdot \(|\cdot ADJ bracis-19048 111 31 |_1\ |_1\ NOUN bracis-19048 111 32 ) ) PUNCT bracis-19048 111 33 denoting denote VERB bracis-19048 111 34 the the DET bracis-19048 111 35 \(\ell \(\ell NOUN bracis-19048 112 1 _ _ PRON bracis-19048 112 2 1\ 1\ NUM bracis-19048 112 3 ) ) PUNCT bracis-19048 112 4 norm norm NOUN bracis-19048 112 5 of of ADP bracis-19048 112 6 a a DET bracis-19048 112 7 vector vector NOUN bracis-19048 112 8 . . PUNCT bracis-19048 113 1 calinski calinski PROPN bracis-19048 113 2 - - PUNCT bracis-19048 113 3 harabasz harabasz NOUN bracis-19048 113 4 index index NOUN bracis-19048 113 5 ( ( PUNCT bracis-19048 113 6 ch ch NOUN bracis-19048 113 7 ) ) PUNCT bracis-19048 113 8     SPACE bracis-19048 114 1 [ [ X bracis-19048 114 2 8 8 NUM bracis-19048 114 3 ] ] SYM bracis-19048 114 4 : : PUNCT bracis-19048 114 5 $ $ SYM bracis-19048 114 6 $ $ SYM bracis-19048 114 7 \begin{aligned \begin{aligne VERB bracis-19048 114 8 } } PUNCT bracis-19048 114 9 ch ch NOUN bracis-19048 114 10 = = SYM bracis-19048 114 11 \frac{\left \frac{\left NOUN bracis-19048 114 12 ( ( PUNCT bracis-19048 114 13 n n CCONJ bracis-19048 114 14 - - PUNCT bracis-19048 114 15 k\right k\right NOUN bracis-19048 114 16 ) ) PUNCT bracis-19048 114 17 bgss}{\left bgss}{\left PROPN bracis-19048 114 18 ( ( PUNCT bracis-19048 114 19 k-1\right k-1\right NOUN bracis-19048 114 20 ) ) PUNCT bracis-19048 114 21 wgss wgss NOUN bracis-19048 114 22 } } PUNCT bracis-19048 114 23 . . PUNCT bracis-19048 115 1 \end{aligned}$$ \end{aligned}$$ PROPN bracis-19048 115 2 ( ( PUNCT bracis-19048 115 3 16 16 NUM bracis-19048 115 4 ) ) PUNCT bracis-19048 115 5 xie xie PROPN bracis-19048 115 6 - - PUNCT bracis-19048 115 7 beni beni PROPN bracis-19048 115 8 index index NOUN bracis-19048 115 9 ( ( PUNCT bracis-19048 115 10 xb xb PROPN bracis-19048 115 11 ) ) PUNCT bracis-19048 115 12     SPACE bracis-19048 116 1 [ [ X bracis-19048 116 2 38 38 NUM bracis-19048 116 3 ] ] SYM bracis-19048 116 4 : : PUNCT bracis-19048 116 5 $ $ SYM bracis-19048 116 6 $ $ SYM bracis-19048 116 7 \begin{aligned \begin{aligne VERB bracis-19048 116 8 } } PUNCT bracis-19048 116 9 xb xb X bracis-19048 117 1 = = PUNCT bracis-19048 117 2 \frac{1}{n}\frac{wgss}{\min \frac{1}{n}\frac{wgss}{\min PROPN bracis-19048 117 3 \limits \limit NOUN bracis-19048 117 4 _ _ PRON bracis-19048 117 5 { { PUNCT bracis-19048 117 6 i\ne i\ne PROPN bracis-19048 117 7 j}\delta j}\delta PROPN bracis-19048 117 8 _ _ PUNCT bracis-19048 117 9 { { PUNCT bracis-19048 117 10 ij ij NOUN bracis-19048 117 11 } } PUNCT bracis-19048 117 12 } } PUNCT bracis-19048 117 13 . . PUNCT bracis-19048 118 1 \end{aligned}$$ \end{aligned}$$ PROPN bracis-19048 118 2 ( ( PUNCT bracis-19048 118 3 17 17 NUM bracis-19048 118 4 ) ) PUNCT bracis-19048 118 5 normalized normalize VERB bracis-19048 118 6 relative relative ADJ bracis-19048 118 7 entropy entropy NOUN bracis-19048 118 8 ( ( PUNCT bracis-19048 118 9 nre nre NOUN bracis-19048 118 10 ) ) PUNCT bracis-19048 118 11     SPACE bracis-19048 119 1 [ [ X bracis-19048 119 2 26 26 NUM bracis-19048 119 3 ] ] X bracis-19048 119 4 : : PUNCT bracis-19048 119 5 $ $ SYM bracis-19048 119 6 $ $ SYM bracis-19048 119 7 \begin{aligned \begin{aligne VERB bracis-19048 119 8 } } PUNCT bracis-19048 119 9 nre nre NOUN bracis-19048 119 10 = = SYM bracis-19048 119 11 \sum \sum PROPN bracis-19048 119 12 _ _ PUNCT bracis-19048 119 13 { { PUNCT bracis-19048 119 14 k=1}^k\frac{n_k}{n}\log k=1}^k\frac{n_k}{n}\log X bracis-19048 119 15 _ _ NOUN bracis-19048 119 16 2\left 2\left NUM bracis-19048 119 17 ( ( PUNCT bracis-19048 119 18 \frac{n_k}{n}\right \frac{n_k}{n}\right PROPN bracis-19048 119 19 ) ) PUNCT bracis-19048 119 20 . . PUNCT bracis-19048 120 1 \end{aligned}$$ \end{aligned}$$ PROPN bracis-19048 120 2 ( ( PUNCT bracis-19048 120 3 18 18 NUM bracis-19048 120 4 ) ) PUNCT bracis-19048 120 5 c c NOUN bracis-19048 120 6 index index NOUN bracis-19048 120 7 ( ( PUNCT bracis-19048 120 8 c c NOUN bracis-19048 120 9 ) ) PUNCT bracis-19048 120 10     SPACE bracis-19048 121 1 [ [ X bracis-19048 121 2 21 21 NUM bracis-19048 121 3 ] ] X bracis-19048 121 4 : : PUNCT bracis-19048 121 5 let let AUX bracis-19048 121 6 \(s_{min}\ \(s_{min}\ PROPN bracis-19048 121 7 ) ) PUNCT bracis-19048 121 8 denote denote VERB bracis-19048 121 9 the the DET bracis-19048 121 10 sum sum NOUN bracis-19048 121 11 of of ADP bracis-19048 121 12 the the DET bracis-19048 121 13 \(n_w\ \(n_w\ NOUN bracis-19048 121 14 ) ) PUNCT bracis-19048 121 15 smallest small ADJ bracis-19048 121 16 elements element NOUN bracis-19048 121 17 of of ADP bracis-19048 121 18 \(\mathbf \(\mathbf NOUN bracis-19048 121 19 { { PUNCT bracis-19048 121 20 d}^\bullet d}^\bullet PROPN bracis-19048 121 21 \ \ PROPN bracis-19048 121 22 ) ) PUNCT bracis-19048 121 23 and and CCONJ bracis-19048 121 24 \(s_{max}\ \(s_{max}\ NOUN bracis-19048 121 25 ) ) PUNCT bracis-19048 121 26 the the DET bracis-19048 121 27 sum sum NOUN bracis-19048 121 28 of of ADP bracis-19048 121 29 its its PRON bracis-19048 121 30 \(n_w\ \(n_w\ X bracis-19048 121 31 ) ) PUNCT bracis-19048 121 32 largest large ADJ bracis-19048 121 33 elements element NOUN bracis-19048 121 34 . . PUNCT bracis-19048 122 1 then then ADV bracis-19048 122 2 : : PUNCT bracis-19048 122 3 $ $ SYM bracis-19048 122 4 $ $ SYM bracis-19048 122 5 \begin{aligned \begin{aligne VERB bracis-19048 122 6 } } PUNCT bracis-19048 122 7 c c NOUN bracis-19048 123 1 = = PUNCT bracis-19048 123 2 \frac{|\mathbf \frac{|\mathbf PUNCT bracis-19048 123 3 { { PUNCT bracis-19048 123 4 d}^+|_1 d}^+|_1 PROPN bracis-19048 123 5 s_{min}}{s_{max s_{min}}{s_{max NOUN bracis-19048 123 6 } } PUNCT bracis-19048 123 7 s_{min s_{min NOUN bracis-19048 123 8 } } PUNCT bracis-19048 123 9 } } PUNCT bracis-19048 123 10 . . PUNCT bracis-19048 124 1 \end{aligned}$$ \end{aligned}$$ PROPN bracis-19048 124 2 ( ( PUNCT bracis-19048 124 3 19 19 NUM bracis-19048 124 4 ) ) PUNCT bracis-19048 124 5 mean mean NOUN bracis-19048 124 6 of of ADP bracis-19048 124 7 distances distance NOUN bracis-19048 124 8 to to ADP bracis-19048 124 9 cluster cluster NOUN bracis-19048 124 10 centroids centroid NOUN bracis-19048 124 11 ( ( PUNCT bracis-19048 124 12 cm cm NOUN bracis-19048 124 13 ): ): PUNCT bracis-19048 124 14 $ $ SYM bracis-19048 124 15 $ $ SYM bracis-19048 124 16 \begin{aligned \begin{aligne VERB bracis-19048 124 17 } } PUNCT bracis-19048 124 18 cm cm NOUN bracis-19048 124 19 = = NOUN bracis-19048 124 20 \frac{1}{k}\sum \frac{1}{k}\sum NOUN bracis-19048 124 21 _ _ PRON bracis-19048 124 22 { { PUNCT bracis-19048 124 23 k=1}^k\sum k=1}^k\sum ADP bracis-19048 124 24 _ _ PRON bracis-19048 124 25 { { PUNCT bracis-19048 124 26 \forall \forall DET bracis-19048 124 27 \mathbf \mathbf PROPN bracis-19048 124 28 { { PUNCT bracis-19048 124 29 x}_i\in x}_i\in PROPN bracis-19048 124 30 x_k}dist(\mathbf x_k}dist(\mathbf PROPN bracis-19048 124 31 { { PUNCT bracis-19048 124 32 x}_i,\mathbf x}_i,\mathbf NOUN bracis-19048 124 33 { { PUNCT bracis-19048 124 34 \bar{x}}_k \bar{x}}_k PROPN bracis-19048 124 35 ) ) PUNCT bracis-19048 124 36 \end{aligned}$$ \end{aligned}$$ X bracis-19048 124 37 ( ( PUNCT bracis-19048 124 38 20 20 NUM bracis-19048 124 39 ) ) PUNCT bracis-19048 124 40 connectivity connectivity NOUN bracis-19048 124 41 ( ( PUNCT bracis-19048 124 42 cn cn NOUN bracis-19048 124 43 ) ) PUNCT bracis-19048 124 44     SPACE bracis-19048 125 1 [ [ X bracis-19048 125 2 7 7 NUM bracis-19048 125 3 , , PUNCT bracis-19048 125 4 19 19 NUM bracis-19048 125 5 ] ] NOUN bracis-19048 125 6 : : PUNCT bracis-19048 125 7 $ $ SYM bracis-19048 125 8 $ $ SYM bracis-19048 125 9 \begin{aligned \begin{aligne VERB bracis-19048 125 10 } } PUNCT bracis-19048 125 11 cn cn NOUN bracis-19048 125 12 = = SYM bracis-19048 125 13 \sum \sum NOUN bracis-19048 125 14 _ _ PUNCT bracis-19048 125 15 { { PUNCT bracis-19048 125 16 i=1}^n\sum i=1}^n\sum NOUN bracis-19048 125 17 _ _ PRON bracis-19048 125 18 { { PUNCT bracis-19048 125 19 j=1}^l\mathcal j=1}^l\mathcal X bracis-19048 125 20 { { PUNCT bracis-19048 125 21 i}\left i}\left NOUN bracis-19048 125 22 ( ( PUNCT bracis-19048 125 23 \mathbf \mathbf PROPN bracis-19048 125 24 { { PUNCT bracis-19048 125 25 x}_i,\eta x}_i,\eta PROPN bracis-19048 125 26 _ _ NOUN bracis-19048 125 27 j\left j\left PROPN bracis-19048 125 28 ( ( PUNCT bracis-19048 125 29 \mathbf \mathbf PROPN bracis-19048 125 30 { { PUNCT bracis-19048 125 31 x}_i\right x}_i\right PROPN bracis-19048 125 32 ) ) PUNCT bracis-19048 126 1 \right \right NOUN bracis-19048 126 2 ) ) PUNCT bracis-19048 126 3 , , PUNCT bracis-19048 126 4 \end{aligned}$$ \end{aligned}$$ X bracis-19048 126 5 ( ( PUNCT bracis-19048 126 6 21 21 NUM bracis-19048 126 7 ) ) PUNCT bracis-19048 127 1 where where SCONJ bracis-19048 127 2 \(\eta \(\eta NOUN bracis-19048 127 3 _ _ NOUN bracis-19048 127 4 j\left j\left NOUN bracis-19048 127 5 ( ( PUNCT bracis-19048 127 6 \mathbf \mathbf PROPN bracis-19048 127 7 { { PUNCT bracis-19048 127 8 x}_i\right x}_i\right PROPN bracis-19048 127 9 ) ) PUNCT bracis-19048 127 10 \ \ NOUN bracis-19048 127 11 ) ) PUNCT bracis-19048 127 12 denotes denote VERB bracis-19048 127 13 the the DET bracis-19048 127 14 j j PROPN bracis-19048 127 15 - - PUNCT bracis-19048 127 16 th th X bracis-19048 127 17 nearest near ADJ bracis-19048 127 18 - - PUNCT bracis-19048 127 19 neighbor neighbor NOUN bracis-19048 127 20 to to PART bracis-19048 127 21 point point VERB bracis-19048 127 22 \(\mathbf \(\mathbf VERB bracis-19048 127 23 { { PUNCT bracis-19048 127 24 x}_i\ x}_i\ NOUN bracis-19048 127 25 ) ) PUNCT bracis-19048 127 26 , , PUNCT bracis-19048 127 27 and and CCONJ bracis-19048 127 28 \(\mathcal \(\mathcal ADJ bracis-19048 127 29 { { PUNCT bracis-19048 127 30 i}\left i}\left NOUN bracis-19048 127 31 ( ( PUNCT bracis-19048 127 32 \mathbf \mathbf PROPN bracis-19048 127 33 { { PUNCT bracis-19048 127 34 x}_i,\eta x}_i,\eta PROPN bracis-19048 127 35 _ _ NOUN bracis-19048 127 36 j\left j\left PROPN bracis-19048 127 37 ( ( PUNCT bracis-19048 127 38 \mathbf \mathbf PROPN bracis-19048 127 39 { { PUNCT bracis-19048 127 40 x}_i\right x}_i\right PROPN bracis-19048 127 41 ) ) PUNCT bracis-19048 127 42 \right \right NOUN bracis-19048 127 43 ) ) PUNCT bracis-19048 127 44 \ \ NOUN bracis-19048 127 45 ) ) PUNCT bracis-19048 127 46 is be AUX bracis-19048 127 47 an an DET bracis-19048 127 48 indicator indicator NOUN bracis-19048 127 49 function function NOUN bracis-19048 127 50 that that PRON bracis-19048 127 51 receives receive VERB bracis-19048 127 52 the the DET bracis-19048 127 53 value value NOUN bracis-19048 127 54 of of ADP bracis-19048 127 55 1 1 NUM bracis-19048 127 56 / / SYM bracis-19048 127 57 j j NOUN bracis-19048 127 58 if if SCONJ bracis-19048 127 59 \(\mathbf \(\mathbf VERB bracis-19048 127 60 { { PUNCT bracis-19048 127 61 x}_i\ x}_i\ NOUN bracis-19048 127 62 ) ) PUNCT bracis-19048 127 63 and and CCONJ bracis-19048 127 64 \(\eta \(\eta X bracis-19048 127 65 _ _ NOUN bracis-19048 127 66 j\left j\left NOUN bracis-19048 127 67 ( ( PUNCT bracis-19048 127 68 \mathbf \mathbf PROPN bracis-19048 127 69 { { PUNCT bracis-19048 127 70 x}_i\right x}_i\right PROPN bracis-19048 127 71 ) ) PUNCT bracis-19048 127 72 \ \ NOUN bracis-19048 127 73 ) ) PUNCT bracis-19048 127 74 belong belong VERB bracis-19048 127 75 to to ADP bracis-19048 127 76 the the DET bracis-19048 127 77 same same ADJ bracis-19048 127 78 cluster cluster NOUN bracis-19048 127 79 , , PUNCT bracis-19048 127 80 zero zero NUM bracis-19048 127 81 otherwise otherwise ADV bracis-19048 127 82 . . PUNCT bracis-19048 128 1 average average ADJ bracis-19048 128 2 scattering scattering NOUN bracis-19048 128 3 for for ADP bracis-19048 128 4 clusters cluster NOUN bracis-19048 128 5 ( ( PUNCT bracis-19048 128 6 \(sd_{scat}\ \(sd_{scat}\ NUM bracis-19048 128 7 ) ) PUNCT bracis-19048 128 8 ) ) PUNCT bracis-19048 128 9     SPACE bracis-19048 129 1 [ [ X bracis-19048 129 2 18 18 NUM bracis-19048 129 3 ] ] SYM bracis-19048 129 4 : : PUNCT bracis-19048 129 5 $ $ SYM bracis-19048 129 6 $ $ SYM bracis-19048 129 7 \begin{aligned \begin{aligne VERB bracis-19048 129 8 } } PUNCT bracis-19048 129 9 sd_{scat sd_{scat PROPN bracis-19048 129 10 } } PUNCT bracis-19048 129 11 = = PUNCT bracis-19048 129 12 \frac{\sum \frac{\sum NOUN bracis-19048 129 13 \limits \limits ADP bracis-19048 129 14 _ _ PRON bracis-19048 129 15 { { PUNCT bracis-19048 129 16 k=1}^k k=1}^k PROPN bracis-19048 129 17 \left| \left| NOUN bracis-19048 129 18 \hat{\boldsymbol{\sigma \hat{\boldsymbol{\sigma X bracis-19048 129 19 } } PUNCT bracis-19048 129 20 } } PUNCT bracis-19048 129 21 ^2_{k}\right| ^2_{k}\right| NUM bracis-19048 129 22 _ _ PUNCT bracis-19048 129 23 2}{k\left| 2}{k\left| NUM bracis-19048 129 24 \hat{\boldsymbol{\sigma \hat{\boldsymbol{\sigma NOUN bracis-19048 129 25 } } PUNCT bracis-19048 129 26 } } PUNCT bracis-19048 129 27 ^2_{\bullet ^2_{\bullet ADP bracis-19048 129 28 } } PUNCT bracis-19048 129 29 \right| \right| X bracis-19048 129 30 _ _ NOUN bracis-19048 129 31 2 2 NUM bracis-19048 129 32 } } PUNCT bracis-19048 129 33 , , PUNCT bracis-19048 129 34 \end{aligned}$$ \end{aligned}$$ X bracis-19048 129 35 ( ( PUNCT bracis-19048 129 36 22 22 NUM bracis-19048 129 37 ) ) PUNCT bracis-19048 129 38 where where SCONJ bracis-19048 129 39 \(\hat{\boldsymbol{\sigma \(\hat{\boldsymbol{\sigma ADP bracis-19048 129 40 } } PUNCT bracis-19048 129 41 } } PUNCT bracis-19048 129 42 ^2_{\bullet ^2_{\bullet ADP bracis-19048 129 43 } } PUNCT bracis-19048 129 44 \ \ NOUN bracis-19048 129 45 ) ) PUNCT bracis-19048 129 46 denotes denote VERB bracis-19048 129 47 the the DET bracis-19048 129 48 vector vector NOUN bracis-19048 129 49 of of ADP bracis-19048 129 50 variance variance NOUN bracis-19048 129 51 estimates estimate NOUN bracis-19048 129 52 for for ADP bracis-19048 129 53 all all DET bracis-19048 129 54 attributes attribute NOUN bracis-19048 129 55 in in ADP bracis-19048 129 56 all all DET bracis-19048 129 57 clusters cluster NOUN bracis-19048 129 58 , , PUNCT bracis-19048 129 59 and and CCONJ bracis-19048 129 60 \(\hat{\boldsymbol{\sigma \(\hat{\boldsymbol{\sigma ADP bracis-19048 129 61 } } PUNCT bracis-19048 129 62 } } PUNCT bracis-19048 129 63 ^2_{k}\ ^2_{k}\ PROPN bracis-19048 129 64 ) ) PUNCT bracis-19048 129 65 is be AUX bracis-19048 129 66 the the DET bracis-19048 129 67 vector vector NOUN bracis-19048 129 68 of of ADP bracis-19048 129 69 variance variance NOUN bracis-19048 129 70 estimates estimate NOUN bracis-19048 129 71 for for ADP bracis-19048 129 72 all all DET bracis-19048 129 73 attributes attribute NOUN bracis-19048 129 74 considering consider VERB bracis-19048 129 75 only only ADV bracis-19048 129 76 the the DET bracis-19048 129 77 observations observation NOUN bracis-19048 129 78 in in ADP bracis-19048 129 79 the the DET bracis-19048 129 80 k k NOUN bracis-19048 129 81 - - PUNCT bracis-19048 129 82 th th VERB bracis-19048 129 83 cluster cluster NOUN bracis-19048 129 84 . . PUNCT bracis-19048 130 1 total total ADJ bracis-19048 130 2 separation separation NOUN bracis-19048 130 3 between between ADP bracis-19048 130 4 clusters cluster NOUN bracis-19048 130 5 ( ( PUNCT bracis-19048 130 6 \(sd_{dis}\ \(sd_{dis}\ PROPN bracis-19048 130 7 ) ) PUNCT bracis-19048 130 8 ) ) PUNCT bracis-19048 130 9     SPACE bracis-19048 131 1 [ [ X bracis-19048 131 2 18 18 NUM bracis-19048 131 3 ] ] SYM bracis-19048 131 4 : : PUNCT bracis-19048 131 5 $ $ SYM bracis-19048 131 6 $ $ SYM bracis-19048 131 7 \begin{aligned \begin{aligne VERB bracis-19048 131 8 } } PUNCT bracis-19048 131 9 sd_{dis sd_{di NOUN bracis-19048 131 10 } } PUNCT bracis-19048 131 11 = = PUNCT bracis-19048 131 12 \frac{\kappa \frac{\kappa X bracis-19048 131 13 _ _ PUNCT bracis-19048 131 14 { { PUNCT bracis-19048 131 15 max}}{\kappa max}}{\kappa ADJ bracis-19048 131 16 _ _ PUNCT bracis-19048 131 17 { { PUNCT bracis-19048 131 18 min}}\sum min}}\sum NOUN bracis-19048 131 19 \limits \limit NOUN bracis-19048 131 20 _ _ PRON bracis-19048 131 21 { { PUNCT bracis-19048 131 22 k=1}^{k}\left k=1}^{k}\left PROPN bracis-19048 131 23 ( ( PUNCT bracis-19048 131 24 \sum \sum NOUN bracis-19048 131 25 _ _ PUNCT bracis-19048 131 26 { { PUNCT bracis-19048 131 27 \begin{array}{c \begin{array}{c PROPN bracis-19048 131 28 } } PUNCT bracis-19048 131 29 k^\prime k^\prime NOUN bracis-19048 131 30 = = SYM bracis-19048 131 31 1\\ 1\\ NUM bracis-19048 131 32 k^\prime k^\prime PROPN bracis-19048 131 33 \ne \ne PROPN bracis-19048 131 34 k k PROPN bracis-19048 131 35 \end{array}}^k \end{array}}^k PUNCT bracis-19048 131 36 dist\left dist\left VERB bracis-19048 131 37 ( ( PUNCT bracis-19048 131 38 \bar{\mathbf \bar{\mathbf PROPN bracis-19048 131 39 { { PUNCT bracis-19048 131 40 x}}_k,\bar{\mathbf x}}_k,\bar{\mathbf PROPN bracis-19048 131 41 { { PUNCT bracis-19048 131 42 x}}_{k^\prime x}}_{k^\prime PROPN bracis-19048 131 43 } } PUNCT bracis-19048 131 44 \right \right PROPN bracis-19048 131 45 ) ) PUNCT bracis-19048 131 46 \right \right NOUN bracis-19048 131 47 ) ) PUNCT bracis-19048 131 48 ^{-1 ^{-1 NOUN bracis-19048 131 49 } } PUNCT bracis-19048 131 50 \end{aligned}$$ \end{aligned}$$ X bracis-19048 131 51 ( ( PUNCT bracis-19048 131 52 23 23 NUM bracis-19048 131 53 ) ) PUNCT bracis-19048 131 54 where where SCONJ bracis-19048 131 55 : : PUNCT bracis-19048 131 56 $ $ SYM bracis-19048 131 57 $ $ SYM bracis-19048 131 58 \begin{aligned \begin{aligne VERB bracis-19048 131 59 } } PUNCT bracis-19048 131 60 \kappa \kappa VERB bracis-19048 131 61 _ _ PRON bracis-19048 131 62 { { PUNCT bracis-19048 131 63 max max PROPN bracis-19048 131 64 } } PUNCT bracis-19048 131 65 = = SYM bracis-19048 131 66 \max \max NOUN bracis-19048 131 67 \limits \limit VERB bracis-19048 131 68 _ _ PRON bracis-19048 131 69 { { PUNCT bracis-19048 131 70 k^\prime k^\prime PROPN bracis-19048 131 71 \ne \ne PROPN bracis-19048 131 72 k k PROPN bracis-19048 131 73 } } PUNCT bracis-19048 131 74 dist\left dist\left VERB bracis-19048 131 75 ( ( PUNCT bracis-19048 131 76 \bar{\mathbf \bar{\mathbf PROPN bracis-19048 131 77 { { PUNCT bracis-19048 131 78 x}}_k,\bar{\mathbf x}}_k,\bar{\mathbf PROPN bracis-19048 131 79 { { PUNCT bracis-19048 131 80 x}}_{k^\prime x}}_{k^\prime PROPN bracis-19048 131 81 } } PUNCT bracis-19048 131 82 \right \right PROPN bracis-19048 131 83 ) ) PUNCT bracis-19048 131 84 \quad \quad PROPN bracis-19048 131 85 \mathrm \mathrm PROPN bracis-19048 131 86 { { PUNCT bracis-19048 131 87 and}\quad and}\quad PUNCT bracis-19048 131 88 \kappa \kappa VERB bracis-19048 131 89 _ _ PUNCT bracis-19048 131 90 { { PUNCT bracis-19048 131 91 min min NOUN bracis-19048 131 92 } } PUNCT bracis-19048 131 93 = = NOUN bracis-19048 131 94 \min \min NOUN bracis-19048 131 95 \limits \limit NOUN bracis-19048 131 96 _ _ PRON bracis-19048 131 97 { { PUNCT bracis-19048 131 98 k^\prime k^\prime PROPN bracis-19048 131 99 \ne \ne PROPN bracis-19048 131 100 k k PROPN bracis-19048 131 101 } } PUNCT bracis-19048 131 102 dist\left dist\left VERB bracis-19048 131 103 ( ( PUNCT bracis-19048 131 104 \bar{\mathbf \bar{\mathbf PROPN bracis-19048 131 105 { { PUNCT bracis-19048 131 106 x}}_k,\bar{\mathbf x}}_k,\bar{\mathbf PROPN bracis-19048 131 107 { { PUNCT bracis-19048 131 108 x}}_{k^\prime x}}_{k^\prime PROPN bracis-19048 131 109 } } PUNCT bracis-19048 131 110 \right \right PROPN bracis-19048 131 111 ) ) PUNCT bracis-19048 131 112 . . PUNCT bracis-19048 132 1 \end{aligned}$$ \end{aligned}$$ PROPN bracis-19048 132 2 ( ( PUNCT bracis-19048 132 3 24 24 NUM bracis-19048 132 4 ) ) PUNCT bracis-19048 132 5 akaike akaike ADJ bracis-19048 132 6 ’s ’s PART bracis-19048 132 7 information information NOUN bracis-19048 132 8 criterion criterion NOUN bracis-19048 132 9 ( ( PUNCT bracis-19048 132 10 aic aic PROPN bracis-19048 132 11 ) ) PUNCT bracis-19048 132 12     SPACE bracis-19048 133 1 [ [ X bracis-19048 133 2 33 33 NUM bracis-19048 133 3 ] ] X bracis-19048 133 4 : : PUNCT bracis-19048 133 5 the the DET bracis-19048 133 6 aic aic PROPN bracis-19048 133 7 of of ADP bracis-19048 133 8 a a DET bracis-19048 133 9 first first ADJ bracis-19048 133 10 - - PUNCT bracis-19048 133 11 order order NOUN bracis-19048 133 12 multiple multiple ADJ bracis-19048 133 13 linear linear ADJ bracis-19048 133 14 regression regression NOUN bracis-19048 133 15 model model NOUN bracis-19048 133 16 of of ADP bracis-19048 133 17 class class NOUN bracis-19048 133 18 labels label NOUN bracis-19048 133 19 on on ADP bracis-19048 133 20 each each DET bracis-19048 133 21 attribute attribute NOUN bracis-19048 133 22 of of ADP bracis-19048 133 23 the the DET bracis-19048 133 24 dataset dataset NOUN bracis-19048 133 25 . . PUNCT bracis-19048 134 1 the the DET bracis-19048 134 2 linear linear ADJ bracis-19048 134 3 model model NOUN bracis-19048 134 4 is be AUX bracis-19048 134 5 given give VERB bracis-19048 134 6 by by ADP bracis-19048 134 7 $ $ SYM bracis-19048 134 8 $ $ SYM bracis-19048 134 9 \begin{aligned \begin{aligne VERB bracis-19048 134 10 } } PUNCT bracis-19048 134 11 class(\mathbf class(\mathbf NOUN bracis-19048 134 12 { { PUNCT bracis-19048 134 13 x}_i x}_i NUM bracis-19048 134 14 ) ) PUNCT bracis-19048 135 1 = = SYM bracis-19048 136 1 \hat{\beta \hat{\beta ADJ bracis-19048 136 2 } } PUNCT bracis-19048 136 3 _ _ NOUN bracis-19048 136 4 0 0 PUNCT bracis-19048 137 1 + + CCONJ bracis-19048 137 2 \hat{\boldsymbol{\beta \hat{\boldsymbol{\beta NOUN bracis-19048 137 3 } } PUNCT bracis-19048 137 4 } } PUNCT bracis-19048 137 5 \mathbf \mathbf PROPN bracis-19048 137 6 { { PUNCT bracis-19048 137 7 x}_i x}_i PROPN bracis-19048 137 8 + + PROPN bracis-19048 137 9 e_i e_i AUX bracis-19048 137 10 , , PUNCT bracis-19048 137 11 \end{aligned}$$ \end{aligned}$$ X bracis-19048 137 12 ( ( PUNCT bracis-19048 137 13 25 25 NUM bracis-19048 137 14 ) ) PUNCT bracis-19048 137 15 with with ADP bracis-19048 137 16 \(class(\mathbf \(class(\mathbf PROPN bracis-19048 137 17 { { PUNCT bracis-19048 137 18 x}_i)\ x}_i)\ ADV bracis-19048 137 19 ) ) PUNCT bracis-19048 137 20 denoting denote VERB bracis-19048 137 21 the the DET bracis-19048 137 22 numerically numerically ADV bracis-19048 137 23 - - PUNCT bracis-19048 137 24 encoded encode VERB bracis-19048 137 25 class class NOUN bracis-19048 137 26 of of ADP bracis-19048 137 27 point point NOUN bracis-19048 137 28 \(\mathbf \(\mathbf VERB bracis-19048 137 29 { { PUNCT bracis-19048 137 30 x}_i\in x}_i\in PROPN bracis-19048 137 31 { { PUNCT bracis-19048 137 32 x}\ x}\ PROPN bracis-19048 137 33 ) ) PUNCT bracis-19048 137 34 , , PUNCT bracis-19048 137 35 \(\hat{\beta \(\hat{\beta PROPN bracis-19048 137 36 } } PUNCT bracis-19048 137 37 _ _ NOUN bracis-19048 137 38 0\ 0\ NUM bracis-19048 137 39 ) ) PUNCT bracis-19048 137 40 and and CCONJ bracis-19048 137 41 \(~\hat{\boldsymbol{\beta \(~\hat{\boldsymbol{\beta PROPN bracis-19048 137 42 } } PUNCT bracis-19048 137 43 } } PUNCT bracis-19048 137 44 \ \ NOUN bracis-19048 137 45 ) ) PUNCT bracis-19048 137 46 representing represent VERB bracis-19048 137 47 the the DET bracis-19048 137 48 fitted fit VERB bracis-19048 137 49 coefficients coefficient NOUN bracis-19048 137 50 of of ADP bracis-19048 137 51 the the DET bracis-19048 137 52 model model NOUN bracis-19048 137 53 , , PUNCT bracis-19048 137 54 and and CCONJ bracis-19048 137 55 \(e_i\ \(e_i\ NOUN bracis-19048 137 56 ) ) PUNCT bracis-19048 137 57 the the DET bracis-19048 137 58 residual residual ADJ bracis-19048 137 59 related relate VERB bracis-19048 137 60 to to ADP bracis-19048 137 61 the the DET bracis-19048 137 62 i i PROPN bracis-19048 137 63 - - PUNCT bracis-19048 137 64 th th VERB bracis-19048 137 65 observation observation NOUN bracis-19048 137 66 in in ADP bracis-19048 137 67 the the DET bracis-19048 137 68 dataset dataset NOUN bracis-19048 137 69 . . PUNCT bracis-19048 138 1 the the DET bracis-19048 138 2 aic aic PROPN bracis-19048 138 3 of of ADP bracis-19048 138 4 the the DET bracis-19048 138 5 model model NOUN bracis-19048 138 6 is be AUX bracis-19048 138 7 then then ADV bracis-19048 138 8 given give VERB bracis-19048 138 9 as as ADP bracis-19048 138 10 : : PUNCT bracis-19048 138 11 $ $ SYM bracis-19048 138 12 $ $ SYM bracis-19048 138 13 \begin{aligned \begin{aligne VERB bracis-19048 138 14 } } PUNCT bracis-19048 138 15 aic aic PROPN bracis-19048 138 16 = = SYM bracis-19048 138 17 -2\ln -2\ln X bracis-19048 138 18 ( ( PUNCT bracis-19048 138 19 l l NOUN bracis-19048 138 20 ) ) PUNCT bracis-19048 138 21 + + CCONJ bracis-19048 138 22 2q 2q NUM bracis-19048 138 23 \end{aligned}$$ \end{aligned}$$ X bracis-19048 138 24 ( ( PUNCT bracis-19048 138 25 26 26 NUM bracis-19048 138 26 ) ) PUNCT bracis-19048 138 27 where where SCONJ bracis-19048 138 28 \(\ln \(\ln PROPN bracis-19048 138 29 { { PUNCT bracis-19048 138 30 l}\ l}\ PROPN bracis-19048 138 31 ) ) PUNCT bracis-19048 138 32 is be AUX bracis-19048 138 33 the the DET bracis-19048 138 34 log log NOUN bracis-19048 138 35 - - PUNCT bracis-19048 138 36 likelihood likelihood NOUN bracis-19048 138 37 value value NOUN bracis-19048 138 38 estimated estimate VERB bracis-19048 138 39 for for ADP bracis-19048 138 40 the the DET bracis-19048 138 41 model model NOUN bracis-19048 138 42 ( ( PUNCT bracis-19048 138 43 25 25 NUM bracis-19048 138 44 ) ) PUNCT bracis-19048 138 45 . . PUNCT bracis-19048 139 1 bayesian bayesian NOUN bracis-19048 139 2 information information NOUN bracis-19048 139 3 criterion criterion NOUN bracis-19048 139 4 ( ( PUNCT bracis-19048 139 5 bic bic NOUN bracis-19048 139 6 ) ) PUNCT bracis-19048 139 7     SPACE bracis-19048 140 1 [ [ X bracis-19048 140 2 33 33 NUM bracis-19048 140 3 ] ] X bracis-19048 140 4 : : PUNCT bracis-19048 140 5 the the DET bracis-19048 140 6 bic bic NOUN bracis-19048 140 7 of of ADP bracis-19048 140 8 the the DET bracis-19048 140 9 regression regression NOUN bracis-19048 140 10 model model NOUN bracis-19048 140 11 described describe VERB bracis-19048 140 12 in in ADP bracis-19048 140 13 ( ( PUNCT bracis-19048 140 14 25 25 NUM bracis-19048 140 15 ) ) PUNCT bracis-19048 140 16 , , PUNCT bracis-19048 140 17 which which PRON bracis-19048 140 18 is be AUX bracis-19048 140 19 calculated calculate VERB bracis-19048 140 20 as as ADP bracis-19048 140 21 : : PUNCT bracis-19048 140 22 $ $ SYM bracis-19048 140 23 $ $ SYM bracis-19048 140 24 \begin{aligned \begin{aligne VERB bracis-19048 140 25 } } PUNCT bracis-19048 140 26 bic bic NOUN bracis-19048 140 27 = = SYM bracis-19048 140 28 -2\ln -2\ln X bracis-19048 140 29 ( ( PUNCT bracis-19048 140 30 l l NOUN bracis-19048 140 31 ) ) PUNCT bracis-19048 140 32 + + CCONJ bracis-19048 140 33 q\ln q\ln NOUN bracis-19048 140 34 ( ( PUNCT bracis-19048 140 35 n n CCONJ bracis-19048 140 36 ) ) PUNCT bracis-19048 140 37 \end{aligned}$$ \end{aligned}$$ X bracis-19048 140 38 ( ( PUNCT bracis-19048 140 39 27 27 NUM bracis-19048 140 40 ) ) SYM bracis-19048 140 41 3 3 NUM bracis-19048 140 42 methodology methodology NOUN bracis-19048 140 43 this this DET bracis-19048 140 44 works work VERB bracis-19048 140 45 aims aim VERB bracis-19048 140 46 to to PART bracis-19048 140 47 investigate investigate VERB bracis-19048 140 48 the the DET bracis-19048 140 49 use use NOUN bracis-19048 140 50 of of ADP bracis-19048 140 51 clustering clustering ADJ bracis-19048 140 52 measures measure NOUN bracis-19048 140 53 as as ADP bracis-19048 140 54 meta meta NOUN bracis-19048 140 55 - - PUNCT bracis-19048 140 56 features feature NOUN bracis-19048 140 57 for for ADP bracis-19048 140 58 learning learn VERB bracis-19048 140 59 a a DET bracis-19048 140 60 recommendation recommendation NOUN bracis-19048 140 61 system system NOUN bracis-19048 140 62 for for ADP bracis-19048 140 63 classification classification NOUN bracis-19048 140 64 tasks task NOUN bracis-19048 140 65 . . PUNCT bracis-19048 141 1 more more ADV bracis-19048 141 2 specifically specifically ADV bracis-19048 141 3 , , PUNCT bracis-19048 141 4 the the DET bracis-19048 141 5 standard standard NOUN bracis-19048 141 6 and and CCONJ bracis-19048 141 7 clustering cluster VERB bracis-19048 141 8 meta meta NOUN bracis-19048 141 9 - - PUNCT bracis-19048 141 10 features feature NOUN bracis-19048 141 11 are be AUX bracis-19048 141 12 used use VERB bracis-19048 141 13 in in ADP bracis-19048 141 14 the the DET bracis-19048 141 15 mtl mtl PROPN bracis-19048 141 16 setup setup NOUN bracis-19048 141 17 designed design VERB bracis-19048 141 18 to to PART bracis-19048 141 19 predict predict VERB bracis-19048 141 20 the the DET bracis-19048 141 21 accuracy accuracy NOUN bracis-19048 141 22 of of ADP bracis-19048 141 23 some some DET bracis-19048 141 24 popular popular ADJ bracis-19048 141 25 classification classification NOUN bracis-19048 141 26 techniques technique NOUN bracis-19048 141 27 for for ADP bracis-19048 141 28 a a DET bracis-19048 141 29 given give VERB bracis-19048 141 30 dataset dataset NOUN bracis-19048 141 31 . . PUNCT bracis-19048 142 1 the the DET bracis-19048 142 2 objectives objective NOUN bracis-19048 142 3 are be AUX bracis-19048 142 4 : : PUNCT bracis-19048 142 5 ( ( PUNCT bracis-19048 142 6 i i NOUN bracis-19048 142 7 ) ) PUNCT bracis-19048 142 8 to to PART bracis-19048 142 9 determine determine VERB bracis-19048 142 10 whether whether SCONJ bracis-19048 142 11 the the DET bracis-19048 142 12 mtl mtl PROPN bracis-19048 142 13 approach approach NOUN bracis-19048 142 14 results result NOUN bracis-19048 142 15 in in ADP bracis-19048 142 16 an an DET bracis-19048 142 17 improved improved ADJ bracis-19048 142 18 recommendation recommendation NOUN bracis-19048 142 19 system system NOUN bracis-19048 142 20 ; ; PUNCT bracis-19048 142 21 ( ( PUNCT bracis-19048 142 22 ii ii NOUN bracis-19048 142 23 ) ) PUNCT bracis-19048 142 24 to to PART bracis-19048 142 25 investigate investigate VERB bracis-19048 142 26 whether whether SCONJ bracis-19048 142 27 the the DET bracis-19048 142 28 clustering cluster VERB bracis-19048 142 29 meta meta ADJ bracis-19048 142 30 - - PUNCT bracis-19048 142 31 features feature NOUN bracis-19048 142 32 contribute contribute VERB bracis-19048 142 33 to to ADP bracis-19048 142 34 the the DET bracis-19048 142 35 performance performance NOUN bracis-19048 142 36 of of ADP bracis-19048 142 37 this this DET bracis-19048 142 38 recommendation recommendation NOUN bracis-19048 142 39 system system NOUN bracis-19048 142 40 ; ; PUNCT bracis-19048 142 41 and and CCONJ bracis-19048 142 42 ( ( PUNCT bracis-19048 142 43 iii iii X bracis-19048 142 44 ) ) PUNCT bracis-19048 142 45 to to PART bracis-19048 142 46 characterize characterize VERB bracis-19048 142 47 the the DET bracis-19048 142 48 execution execution NOUN bracis-19048 142 49 times time NOUN bracis-19048 142 50 required require VERB bracis-19048 142 51 to to PART bracis-19048 142 52 extract extract VERB bracis-19048 142 53 each each DET bracis-19048 142 54 set set NOUN bracis-19048 142 55 of of ADP bracis-19048 142 56 meta meta NOUN bracis-19048 142 57 - - PUNCT bracis-19048 142 58 features feature NOUN bracis-19048 142 59 . . PUNCT bracis-19048 143 1 to to PART bracis-19048 143 2 train train VERB bracis-19048 143 3 and and CCONJ bracis-19048 143 4 assess assess VERB bracis-19048 143 5 the the DET bracis-19048 143 6 meta meta ADJ bracis-19048 143 7 - - PUNCT bracis-19048 143 8 learner learner NOUN bracis-19048 143 9 , , PUNCT bracis-19048 143 10 a a DET bracis-19048 143 11 meta meta ADJ bracis-19048 143 12 - - PUNCT bracis-19048 143 13 dataset dataset NOUN bracis-19048 143 14 was be AUX bracis-19048 143 15 populated populate VERB bracis-19048 143 16 with with ADP bracis-19048 143 17 the the DET bracis-19048 143 18 meta meta ADJ bracis-19048 143 19 - - PUNCT bracis-19048 143 20 feature feature NOUN bracis-19048 143 21 values value NOUN bracis-19048 143 22 ( ( PUNCT bracis-19048 143 23 both both PRON bracis-19048 143 24 clustering cluster VERB bracis-19048 143 25 and and CCONJ bracis-19048 143 26 standard standard ADJ bracis-19048 143 27 meta meta NOUN bracis-19048 143 28 - - PUNCT bracis-19048 143 29 features feature NOUN bracis-19048 143 30 ) ) PUNCT bracis-19048 143 31 for for ADP bracis-19048 143 32 a a DET bracis-19048 143 33 collection collection NOUN bracis-19048 143 34 of of ADP bracis-19048 143 35 problem problem NOUN bracis-19048 143 36 instances instance NOUN bracis-19048 143 37 , , PUNCT bracis-19048 143 38 labeled label VERB bracis-19048 143 39 with with ADP bracis-19048 143 40 the the DET bracis-19048 143 41 performances performance NOUN bracis-19048 143 42 of of ADP bracis-19048 143 43 known know VERB bracis-19048 143 44 classification classification NOUN bracis-19048 143 45 techniques technique NOUN bracis-19048 143 46 . . PUNCT bracis-19048 144 1 the the DET bracis-19048 144 2 meta meta ADJ bracis-19048 144 3 - - PUNCT bracis-19048 144 4 models model NOUN bracis-19048 144 5 were be AUX bracis-19048 144 6 induced induce VERB bracis-19048 144 7 by by ADP bracis-19048 144 8 regression regression NOUN bracis-19048 144 9 techniques technique NOUN bracis-19048 144 10 and and CCONJ bracis-19048 144 11 evaluated evaluate VERB bracis-19048 144 12 in in ADP bracis-19048 144 13 the the DET bracis-19048 144 14 base base NOUN bracis-19048 144 15 - - PUNCT bracis-19048 144 16 level level NOUN bracis-19048 144 17 , , PUNCT bracis-19048 144 18 measuring measure VERB bracis-19048 144 19 the the DET bracis-19048 144 20 predictive predictive ADJ bracis-19048 144 21 performance performance NOUN bracis-19048 144 22 of of ADP bracis-19048 144 23 the the DET bracis-19048 144 24 classifiers classifier NOUN bracis-19048 144 25 , , PUNCT bracis-19048 144 26 and and CCONJ bracis-19048 144 27 the the PRON bracis-19048 144 28 and and CCONJ bracis-19048 144 29 meta meta ADJ bracis-19048 144 30 - - PUNCT bracis-19048 144 31 level level NOUN bracis-19048 144 32 , , PUNCT bracis-19048 144 33 measuring measure VERB bracis-19048 144 34 the the DET bracis-19048 144 35 impact impact NOUN bracis-19048 144 36 of of ADP bracis-19048 144 37 recommending recommend VERB bracis-19048 144 38 the the DET bracis-19048 144 39 best good ADJ bracis-19048 144 40 one one NUM bracis-19048 144 41 . . PUNCT bracis-19048 145 1 the the DET bracis-19048 145 2 computational computational ADJ bracis-19048 145 3 costs cost NOUN bracis-19048 145 4 for for ADP bracis-19048 145 5 theses thesis NOUN bracis-19048 145 6 processes process NOUN bracis-19048 145 7 was be AUX bracis-19048 145 8 recorded record VERB bracis-19048 145 9 for for ADP bracis-19048 145 10 evaluation evaluation NOUN bracis-19048 145 11 . . PUNCT bracis-19048 146 1 four four NUM bracis-19048 146 2 hundred hundred NUM bracis-19048 146 3 datasets dataset NOUN bracis-19048 146 4 from from ADP bracis-19048 146 5 the the DET bracis-19048 146 6 openml openml NOUN bracis-19048 146 7 repository repository NOUN bracis-19048 146 8     SPACE bracis-19048 146 9 [ [ X bracis-19048 146 10 36 36 NUM bracis-19048 146 11 ] ] PUNCT bracis-19048 146 12 , , PUNCT bracis-19048 146 13 representing represent VERB bracis-19048 146 14 diverse diverse ADJ bracis-19048 146 15 application application NOUN bracis-19048 146 16 contexts context NOUN bracis-19048 146 17 and and CCONJ bracis-19048 146 18 domains domain NOUN bracis-19048 146 19 , , PUNCT bracis-19048 146 20 were be AUX bracis-19048 146 21 used use VERB bracis-19048 146 22 in in ADP bracis-19048 146 23 this this DET bracis-19048 146 24 experiment experiment NOUN bracis-19048 146 25 . . PUNCT bracis-19048 147 1 they they PRON bracis-19048 147 2 were be AUX bracis-19048 147 3 selected select VERB bracis-19048 147 4 considering consider VERB bracis-19048 147 5 a a DET bracis-19048 147 6 maximum maximum ADJ bracis-19048 147 7 number number NOUN bracis-19048 147 8 of of ADP bracis-19048 147 9 10 10 NUM bracis-19048 147 10 , , PUNCT bracis-19048 147 11     SPACE bracis-19048 147 12 000 000 NUM bracis-19048 147 13 observations observation NOUN bracis-19048 147 14 , , PUNCT bracis-19048 147 15 500 500 NUM bracis-19048 147 16 features feature NOUN bracis-19048 147 17 and and CCONJ bracis-19048 147 18 10 10 NUM bracis-19048 147 19 classes class NOUN bracis-19048 147 20 , , PUNCT bracis-19048 147 21 to to PART bracis-19048 147 22 constrain constrain VERB bracis-19048 147 23 the the DET bracis-19048 147 24 computational computational ADJ bracis-19048 147 25 costs cost NOUN bracis-19048 147 26 of of ADP bracis-19048 147 27 the the DET bracis-19048 147 28 process process NOUN bracis-19048 147 29 . . PUNCT bracis-19048 148 1 for for ADP bracis-19048 148 2 each each DET bracis-19048 148 3 dataset dataset NOUN bracis-19048 148 4 , , PUNCT bracis-19048 148 5 both both CCONJ bracis-19048 148 6 standard standard ADJ bracis-19048 148 7 and and CCONJ bracis-19048 148 8 clustering cluster VERB bracis-19048 148 9 meta meta NOUN bracis-19048 148 10 - - PUNCT bracis-19048 148 11 features feature NOUN bracis-19048 148 12 , , PUNCT bracis-19048 148 13 as as SCONJ bracis-19048 148 14 described describe VERB bracis-19048 148 15 in in ADP bracis-19048 148 16 sect sect NOUN bracis-19048 148 17 . . PUNCT bracis-19048 148 18     SPACE bracis-19048 149 1 2.2 2.2 NUM bracis-19048 149 2 , , PUNCT bracis-19048 149 3 were be AUX bracis-19048 149 4 computed compute VERB bracis-19048 149 5 . . PUNCT bracis-19048 150 1 the the DET bracis-19048 150 2 averages average NOUN bracis-19048 150 3 of of ADP bracis-19048 150 4 10 10 NUM bracis-19048 150 5 - - ADJ bracis-19048 150 6 fold fold ADJ bracis-19048 150 7 cross cross ADJ bracis-19048 150 8 - - ADJ bracis-19048 150 9 validated validated ADJ bracis-19048 150 10 predictive predictive ADJ bracis-19048 150 11 accuracies accuracy NOUN bracis-19048 150 12 achieved achieve VERB bracis-19048 150 13 by by ADP bracis-19048 150 14 each each PRON bracis-19048 150 15 of of ADP bracis-19048 150 16 five five NUM bracis-19048 150 17 classification classification NOUN bracis-19048 150 18 techniques technique NOUN bracis-19048 150 19 , , PUNCT bracis-19048 150 20 for for ADP bracis-19048 150 21 each each DET bracis-19048 150 22 dataset dataset NOUN bracis-19048 150 23 , , PUNCT bracis-19048 150 24 were be AUX bracis-19048 150 25 also also ADV bracis-19048 150 26 calculated calculate VERB bracis-19048 150 27 for for ADP bracis-19048 150 28 labeling label VERB bracis-19048 150 29 the the DET bracis-19048 150 30 meta meta NOUN bracis-19048 150 31 - - PUNCT bracis-19048 150 32 examples example NOUN bracis-19048 150 33 . . PUNCT bracis-19048 151 1 the the DET bracis-19048 151 2 classification classification NOUN bracis-19048 151 3 approaches approach NOUN bracis-19048 151 4 used use VERB bracis-19048 151 5 were be AUX bracis-19048 151 6 : : PUNCT bracis-19048 151 7 c4.5 c4.5 PROPN bracis-19048 151 8 decision decision NOUN bracis-19048 151 9 tree tree NOUN bracis-19048 151 10     SPACE bracis-19048 151 11 [ [ X bracis-19048 151 12 27 27 NUM bracis-19048 151 13 ] ] PUNCT bracis-19048 151 14 with with ADP bracis-19048 151 15 pruning prune VERB bracis-19048 151 16 based base VERB bracis-19048 151 17 on on ADP bracis-19048 151 18 subtree subtree NOUN bracis-19048 151 19 raising raising NOUN bracis-19048 151 20 ; ; PUNCT bracis-19048 151 21 k k X bracis-19048 151 22 - - PUNCT bracis-19048 151 23 nearest near ADJ bracis-19048 151 24 neighbors neighbor NOUN bracis-19048 151 25 ( ( PUNCT bracis-19048 151 26 knn knn PROPN bracis-19048 151 27 ) ) PUNCT bracis-19048 151 28 model model NOUN bracis-19048 151 29     SPACE bracis-19048 152 1 [ [ X bracis-19048 152 2 22 22 NUM bracis-19048 152 3 ] ] PUNCT bracis-19048 152 4 with with ADP bracis-19048 152 5 \(k=3\ \(k=3\ NOUN bracis-19048 152 6 ) ) PUNCT bracis-19048 152 7 ; ; PUNCT bracis-19048 152 8 multilayer multilayer PROPN bracis-19048 152 9 perceptron perceptron PROPN bracis-19048 152 10 ( ( PUNCT bracis-19048 152 11 mlp mlp PROPN bracis-19048 152 12 ) ) PUNCT bracis-19048 152 13     SPACE bracis-19048 153 1 [ [ X bracis-19048 153 2 20 20 NUM bracis-19048 153 3 ] ] PUNCT bracis-19048 153 4 with with ADP bracis-19048 153 5 learning learn VERB bracis-19048 153 6 rate rate NOUN bracis-19048 153 7 of of ADP bracis-19048 153 8 0.3 0.3 NUM bracis-19048 153 9 , , PUNCT bracis-19048 153 10 momentum momentum NOUN bracis-19048 153 11 of of ADP bracis-19048 153 12 0.5 0.5 NUM bracis-19048 153 13 and and CCONJ bracis-19048 153 14 a a DET bracis-19048 153 15 single single ADJ bracis-19048 153 16 hidden hide VERB bracis-19048 153 17 layer layer NOUN bracis-19048 153 18 ; ; PUNCT bracis-19048 153 19 random random ADJ bracis-19048 153 20 forest forest NOUN bracis-19048 153 21 ( ( PUNCT bracis-19048 153 22 rf rf NOUN bracis-19048 153 23 ) ) PUNCT bracis-19048 153 24     SPACE bracis-19048 154 1 [ [ X bracis-19048 154 2 5 5 NUM bracis-19048 154 3 ] ] PUNCT bracis-19048 154 4 with with ADP bracis-19048 154 5 500 500 NUM bracis-19048 154 6 trees tree NOUN bracis-19048 154 7 ; ; PUNCT bracis-19048 154 8 and and CCONJ bracis-19048 154 9 support support VERB bracis-19048 154 10 vector vector NOUN bracis-19048 154 11 machine machine NOUN bracis-19048 154 12 ( ( PUNCT bracis-19048 154 13 svm svm PROPN bracis-19048 154 14 ) ) PUNCT bracis-19048 154 15     SPACE bracis-19048 155 1 [ [ X bracis-19048 155 2 10 10 NUM bracis-19048 155 3 ] ] PUNCT bracis-19048 155 4 with with ADP bracis-19048 155 5 radial radial ADJ bracis-19048 155 6 basis basis NOUN bracis-19048 155 7 kernel kernel NOUN bracis-19048 155 8 . . PUNCT bracis-19048 156 1 these these DET bracis-19048 156 2 hyper hyper ADJ bracis-19048 156 3 - - ADJ bracis-19048 156 4 parameter parameter ADJ bracis-19048 156 5 values value NOUN bracis-19048 156 6 were be AUX bracis-19048 156 7 defined define VERB bracis-19048 156 8 following follow VERB bracis-19048 156 9 the the DET bracis-19048 156 10 standard standard ADJ bracis-19048 156 11 configurations configuration NOUN bracis-19048 156 12 of of ADP bracis-19048 156 13 the the DET bracis-19048 156 14 implementations implementation NOUN bracis-19048 156 15 used use VERB bracis-19048 156 16 . . PUNCT bracis-19048 157 1 this this DET bracis-19048 157 2 process process NOUN bracis-19048 157 3 resulted result VERB bracis-19048 157 4 in in ADP bracis-19048 157 5 a a DET bracis-19048 157 6 meta meta ADJ bracis-19048 157 7 - - PUNCT bracis-19048 157 8 dataset dataset NOUN bracis-19048 157 9 containing contain VERB bracis-19048 157 10 130 130 NUM bracis-19048 157 11 meta meta NOUN bracis-19048 157 12 - - PUNCT bracis-19048 157 13 features feature NOUN bracis-19048 157 14 ( ( PUNCT bracis-19048 157 15 112 112 NUM bracis-19048 157 16 standard standard NOUN bracis-19048 157 17 and and CCONJ bracis-19048 157 18 18 18 NUM bracis-19048 157 19 clustering clustering NOUN bracis-19048 157 20 ) ) PUNCT bracis-19048 157 21 and and CCONJ bracis-19048 157 22 400 400 NUM bracis-19048 157 23 samples sample NOUN bracis-19048 157 24 for for ADP bracis-19048 157 25 each each DET bracis-19048 157 26 classifier classifier NOUN bracis-19048 157 27 , , PUNCT bracis-19048 157 28 labeled label VERB bracis-19048 157 29 by by ADP bracis-19048 157 30 the the DET bracis-19048 157 31 mean mean ADJ bracis-19048 157 32 accuracy accuracy NOUN bracis-19048 157 33 of of ADP bracis-19048 157 34 the the DET bracis-19048 157 35 classification classification NOUN bracis-19048 157 36 method method NOUN bracis-19048 157 37 . . PUNCT bracis-19048 158 1 this this DET bracis-19048 158 2 meta meta ADJ bracis-19048 158 3 - - PUNCT bracis-19048 158 4 dataset dataset NOUN bracis-19048 158 5 was be AUX bracis-19048 158 6 then then ADV bracis-19048 158 7 used use VERB bracis-19048 158 8 to to PART bracis-19048 158 9 train train VERB bracis-19048 158 10 regression regression NOUN bracis-19048 158 11 models model NOUN bracis-19048 158 12 to to PART bracis-19048 158 13 estimate estimate VERB bracis-19048 158 14 the the DET bracis-19048 158 15 expected expect VERB bracis-19048 158 16 accuracy accuracy NOUN bracis-19048 158 17 of of ADP bracis-19048 158 18 each each DET bracis-19048 158 19 classifier classifier NOUN bracis-19048 158 20 , , PUNCT bracis-19048 158 21 as as ADP bracis-19048 158 22 a a DET bracis-19048 158 23 function function NOUN bracis-19048 158 24 of of ADP bracis-19048 158 25 the the DET bracis-19048 158 26 meta meta NOUN bracis-19048 158 27 - - PUNCT bracis-19048 158 28 features feature NOUN bracis-19048 158 29 . . PUNCT bracis-19048 159 1 five five NUM bracis-19048 159 2 regression regression NOUN bracis-19048 159 3 techniques technique NOUN bracis-19048 159 4 known know VERB bracis-19048 159 5 to to PART bracis-19048 159 6 have have VERB bracis-19048 159 7 different different ADJ bracis-19048 159 8 biases bias NOUN bracis-19048 159 9 were be AUX bracis-19048 159 10 tested test VERB bracis-19048 159 11 : : PUNCT bracis-19048 159 12 classification classification NOUN bracis-19048 159 13 and and CCONJ bracis-19048 159 14 regression regression NOUN bracis-19048 159 15 trees tree NOUN bracis-19048 159 16 ( ( PUNCT bracis-19048 159 17 cart cart NOUN bracis-19048 159 18 ) ) PUNCT bracis-19048 159 19     SPACE bracis-19048 160 1 [ [ X bracis-19048 160 2 6 6 NUM bracis-19048 160 3 ] ] PUNCT bracis-19048 160 4 algorithm algorithm NOUN bracis-19048 160 5 with with ADP bracis-19048 160 6 pruning pruning NOUN bracis-19048 160 7 ; ; PUNCT bracis-19048 160 8 distance distance NOUN bracis-19048 160 9 - - PUNCT bracis-19048 160 10 weighted weight VERB bracis-19048 160 11 k k NOUN bracis-19048 160 12 - - PUNCT bracis-19048 160 13 nearest near ADJ bracis-19048 160 14 neighbor neighbor NOUN bracis-19048 160 15 ( ( PUNCT bracis-19048 160 16 dwnn dwnn NOUN bracis-19048 160 17 ) ) PUNCT bracis-19048 160 18     SPACE bracis-19048 161 1 [ [ X bracis-19048 161 2 22 22 NUM bracis-19048 161 3 ] ] PUNCT bracis-19048 161 4 with with ADP bracis-19048 161 5 \(k=3\ \(k=3\ NOUN bracis-19048 161 6 ) ) PUNCT bracis-19048 161 7 and and CCONJ bracis-19048 161 8 gaussian gaussian ADJ bracis-19048 161 9 kernel kernel NOUN bracis-19048 161 10 ; ; PUNCT bracis-19048 161 11 multilayer multilayer PROPN bracis-19048 161 12 perceptron perceptron PROPN bracis-19048 161 13 regression regression PROPN bracis-19048 161 14 ( ( PUNCT bracis-19048 161 15 mlpr mlpr PROPN bracis-19048 161 16 ) ) PUNCT bracis-19048 161 17     SPACE bracis-19048 162 1 [ [ X bracis-19048 162 2 20 20 NUM bracis-19048 162 3 ] ] PUNCT bracis-19048 162 4 with with ADP bracis-19048 162 5 learning learn VERB bracis-19048 162 6 rate rate NOUN bracis-19048 162 7 of of ADP bracis-19048 162 8 0.3 0.3 NUM bracis-19048 162 9 , , PUNCT bracis-19048 162 10 momentum momentum NOUN bracis-19048 162 11 of of ADP bracis-19048 162 12 0.5 0.5 NUM bracis-19048 162 13 and and CCONJ bracis-19048 162 14 a a DET bracis-19048 162 15 single single ADJ bracis-19048 162 16 hidden hide VERB bracis-19048 162 17 layer layer NOUN bracis-19048 162 18 ; ; PUNCT bracis-19048 162 19 random random ADJ bracis-19048 162 20 forest forest NOUN bracis-19048 162 21 regressor regressor NOUN bracis-19048 162 22 ( ( PUNCT bracis-19048 162 23 rfr rfr PROPN bracis-19048 162 24 ) ) PUNCT bracis-19048 162 25     SPACE bracis-19048 163 1 [ [ X bracis-19048 163 2 5 5 NUM bracis-19048 163 3 ] ] PUNCT bracis-19048 163 4 with with ADP bracis-19048 163 5 500 500 NUM bracis-19048 163 6 trees tree NOUN bracis-19048 163 7 ; ; PUNCT bracis-19048 163 8 and and CCONJ bracis-19048 163 9 support support VERB bracis-19048 163 10 vector vector NOUN bracis-19048 163 11 regression regression NOUN bracis-19048 163 12 ( ( PUNCT bracis-19048 163 13 svr svr PROPN bracis-19048 163 14 ) ) PUNCT bracis-19048 163 15     SPACE bracis-19048 164 1 [ [ X bracis-19048 164 2 10 10 NUM bracis-19048 164 3 ] ] PUNCT bracis-19048 164 4 using use VERB bracis-19048 164 5 radial radial ADJ bracis-19048 164 6 basis basis NOUN bracis-19048 164 7 kernel kernel NOUN bracis-19048 164 8 . . PUNCT bracis-19048 165 1 as as ADP bracis-19048 165 2 with with ADP bracis-19048 165 3 the the DET bracis-19048 165 4 classifiers classifier NOUN bracis-19048 165 5 , , PUNCT bracis-19048 165 6 the the DET bracis-19048 165 7 regressor regressor NOUN bracis-19048 165 8 hyper hyper NOUN bracis-19048 165 9 - - NOUN bracis-19048 165 10 parameters parameter NOUN bracis-19048 165 11 were be AUX bracis-19048 165 12 also also ADV bracis-19048 165 13 set set VERB bracis-19048 165 14 as as ADP bracis-19048 165 15 the the DET bracis-19048 165 16 default default NOUN bracis-19048 165 17 values value NOUN bracis-19048 165 18 of of ADP bracis-19048 165 19 the the DET bracis-19048 165 20 implementations implementation NOUN bracis-19048 165 21 used use VERB bracis-19048 165 22 without without ADP bracis-19048 165 23 any any DET bracis-19048 165 24 problem problem NOUN bracis-19048 165 25 - - PUNCT bracis-19048 165 26 specific specific ADJ bracis-19048 165 27 tuning tuning NOUN bracis-19048 165 28 . . PUNCT bracis-19048 166 1 to to PART bracis-19048 166 2 train train VERB bracis-19048 166 3 and and CCONJ bracis-19048 166 4 evaluate evaluate VERB bracis-19048 166 5 the the DET bracis-19048 166 6 regression regression NOUN bracis-19048 166 7 models model NOUN bracis-19048 166 8 , , PUNCT bracis-19048 166 9 10 10 NUM bracis-19048 166 10 - - ADJ bracis-19048 166 11 fold fold ADJ bracis-19048 166 12 cross cross ADJ bracis-19048 166 13 - - ADJ bracis-19048 166 14 validation validation ADJ bracis-19048 166 15 rounds round NOUN bracis-19048 166 16 were be AUX bracis-19048 166 17 also also ADV bracis-19048 166 18 executed execute VERB bracis-19048 166 19 . . PUNCT bracis-19048 167 1 the the DET bracis-19048 167 2 models model NOUN bracis-19048 167 3 obtained obtain VERB bracis-19048 167 4 were be AUX bracis-19048 167 5 evaluated evaluate VERB bracis-19048 167 6 for for ADP bracis-19048 167 7 quality quality NOUN bracis-19048 167 8 , , PUNCT bracis-19048 167 9 considering consider VERB bracis-19048 167 10 a a DET bracis-19048 167 11 comparison comparison NOUN bracis-19048 167 12 to to ADP bracis-19048 167 13 two two NUM bracis-19048 167 14 simple simple ADJ bracis-19048 167 15 baselines baseline NOUN bracis-19048 167 16 : : PUNCT bracis-19048 167 17 random random ADJ bracis-19048 167 18 ( ( PUNCT bracis-19048 167 19 rd rd NOUN bracis-19048 167 20 ) ) PUNCT bracis-19048 167 21 and and CCONJ bracis-19048 167 22 default default NOUN bracis-19048 167 23 ( ( PUNCT bracis-19048 167 24 df df PROPN bracis-19048 167 25 ) ) PUNCT bracis-19048 167 26 . . PUNCT bracis-19048 168 1 the the DET bracis-19048 168 2 rd rd PROPN bracis-19048 168 3 baseline baseline PROPN bracis-19048 168 4 represents represent VERB bracis-19048 168 5 the the DET bracis-19048 168 6 observed observed ADJ bracis-19048 168 7 performance performance NOUN bracis-19048 168 8 of of ADP bracis-19048 168 9 a a DET bracis-19048 168 10 randomly randomly ADV bracis-19048 168 11 chosen choose VERB bracis-19048 168 12 classifier classifier NOUN bracis-19048 168 13 in in ADP bracis-19048 168 14 the the DET bracis-19048 168 15 meta meta ADJ bracis-19048 168 16 - - PUNCT bracis-19048 168 17 base base NOUN bracis-19048 168 18 . . PUNCT bracis-19048 169 1 the the DET bracis-19048 169 2 df df PROPN bracis-19048 169 3 baseline baseline NOUN bracis-19048 169 4 was be AUX bracis-19048 169 5 set set VERB bracis-19048 169 6 as as ADP bracis-19048 169 7 the the DET bracis-19048 169 8 accuracy accuracy NOUN bracis-19048 169 9 of of ADP bracis-19048 169 10 the the DET bracis-19048 169 11 classifier classifier NOUN bracis-19048 169 12 that that PRON bracis-19048 169 13 most most ADV bracis-19048 169 14 often often ADV bracis-19048 169 15 presented present VERB bracis-19048 169 16 the the DET bracis-19048 169 17 best good ADJ bracis-19048 169 18 classification classification NOUN bracis-19048 169 19 performance performance NOUN bracis-19048 169 20 across across ADP bracis-19048 169 21 all all DET bracis-19048 169 22 datasets dataset NOUN bracis-19048 169 23 . . PUNCT bracis-19048 170 1 the the DET bracis-19048 170 2 final final ADJ bracis-19048 170 3 step step NOUN bracis-19048 170 4 is be AUX bracis-19048 170 5 the the DET bracis-19048 170 6 analysis analysis NOUN bracis-19048 170 7 of of ADP bracis-19048 170 8 the the DET bracis-19048 170 9 trade trade NOUN bracis-19048 170 10 - - PUNCT bracis-19048 170 11 off off NOUN bracis-19048 170 12 between between ADP bracis-19048 170 13 the the DET bracis-19048 170 14 computational computational ADJ bracis-19048 170 15 cost cost NOUN bracis-19048 170 16 of of ADP bracis-19048 170 17 each each DET bracis-19048 170 18 set set NOUN bracis-19048 170 19 of of ADP bracis-19048 170 20 meta meta NOUN bracis-19048 170 21 - - PUNCT bracis-19048 170 22 features feature NOUN bracis-19048 170 23 and and CCONJ bracis-19048 170 24 that that PRON bracis-19048 170 25 of of ADP bracis-19048 170 26 evaluating evaluate VERB bracis-19048 170 27 all all DET bracis-19048 170 28 classifiers classifier NOUN bracis-19048 170 29 , , PUNCT bracis-19048 170 30 in in ADP bracis-19048 170 31 a a DET bracis-19048 170 32 cross cross ADJ bracis-19048 170 33 - - ADJ bracis-19048 170 34 validation validation ADJ bracis-19048 170 35 setup setup NOUN bracis-19048 170 36 , , PUNCT bracis-19048 170 37 through through ADP bracis-19048 170 38 direct direct ADJ bracis-19048 170 39 comparison comparison NOUN bracis-19048 170 40 of of ADP bracis-19048 170 41 single single ADJ bracis-19048 170 42 threaded thread VERB bracis-19048 170 43 experiments experiment NOUN bracis-19048 170 44 . . PUNCT bracis-19048 171 1 the the DET bracis-19048 171 2 experiments experiment NOUN bracis-19048 171 3 were be AUX bracis-19048 171 4 done do VERB bracis-19048 171 5 in in ADP bracis-19048 171 6 a a DET bracis-19048 171 7 cluster cluster NOUN bracis-19048 171 8 node node NOUN bracis-19048 171 9 with with ADP bracis-19048 171 10 two two NUM bracis-19048 171 11 intel intel PROPN bracis-19048 171 12 xeon xeon PROPN bracis-19048 171 13 e5 e5 PROPN bracis-19048 171 14 - - PUNCT bracis-19048 171 15 2680v2 2680v2 NUM bracis-19048 171 16 processors processor NOUN bracis-19048 171 17 and and CCONJ bracis-19048 171 18 128 128 NUM bracis-19048 171 19 gb gb PROPN bracis-19048 171 20 ddr3 ddr3 PROPN bracis-19048 171 21 . . PUNCT bracis-19048 172 1 the the DET bracis-19048 172 2 standard standard ADJ bracis-19048 172 3 meta meta NOUN bracis-19048 172 4 - - PUNCT bracis-19048 172 5 features feature NOUN bracis-19048 172 6 , , PUNCT bracis-19048 172 7 as as ADV bracis-19048 172 8 well well ADV bracis-19048 172 9 as as ADP bracis-19048 172 10 some some PRON bracis-19048 172 11 of of ADP bracis-19048 172 12 the the DET bracis-19048 172 13 clustering clustering ADJ bracis-19048 172 14 ones one NOUN bracis-19048 172 15 , , PUNCT bracis-19048 172 16 are be AUX bracis-19048 172 17 provided provide VERB bracis-19048 172 18 by by ADP bracis-19048 172 19 the the DET bracis-19048 172 20 mfe mfe PROPN bracis-19048 172 21 packagefootnote packagefootnote NOUN bracis-19048 172 22 1 1 NUM bracis-19048 172 23 . . X bracis-19048 172 24 4 4 NUM bracis-19048 172 25 experimental experimental ADJ bracis-19048 172 26 results result NOUN bracis-19048 172 27 since since SCONJ bracis-19048 172 28 the the DET bracis-19048 172 29 main main ADJ bracis-19048 172 30 goal goal NOUN bracis-19048 172 31 of of ADP bracis-19048 172 32 algorithm algorithm NOUN bracis-19048 172 33 recommender recommender NOUN bracis-19048 172 34 systems system NOUN bracis-19048 172 35 using use VERB bracis-19048 172 36 mtl mtl PROPN bracis-19048 172 37 is be AUX bracis-19048 172 38 to to PART bracis-19048 172 39 suggest suggest VERB bracis-19048 172 40 , , PUNCT bracis-19048 172 41 within within ADP bracis-19048 172 42 the the DET bracis-19048 172 43 known know VERB bracis-19048 172 44 options option NOUN bracis-19048 172 45 , , PUNCT bracis-19048 172 46 the the DET bracis-19048 172 47 algorithm algorithm NOUN bracis-19048 172 48 with with ADP bracis-19048 172 49 the the DET bracis-19048 172 50 most most ADV bracis-19048 172 51 appropriate appropriate ADJ bracis-19048 172 52 bias bias NOUN bracis-19048 172 53 to to ADP bracis-19048 172 54 a a DET bracis-19048 172 55 particular particular ADJ bracis-19048 172 56 dataset dataset NOUN bracis-19048 172 57     SPACE bracis-19048 173 1 [ [ X bracis-19048 173 2 4 4 NUM bracis-19048 173 3 ] ] PUNCT bracis-19048 173 4 , , PUNCT bracis-19048 173 5 the the DET bracis-19048 173 6 first first ADJ bracis-19048 173 7 analysis analysis NOUN bracis-19048 173 8 of of ADP bracis-19048 173 9 this this DET bracis-19048 173 10 experiment experiment NOUN bracis-19048 173 11 focuses focus VERB bracis-19048 173 12 on on ADP bracis-19048 173 13 such such ADJ bracis-19048 173 14 objective objective NOUN bracis-19048 173 15 . . PUNCT bracis-19048 174 1 figure figure NOUN bracis-19048 174 2     SPACE bracis-19048 174 3 1 1 NUM bracis-19048 174 4 presents present VERB bracis-19048 174 5 the the DET bracis-19048 174 6 results result NOUN bracis-19048 174 7 of of ADP bracis-19048 174 8 two two NUM bracis-19048 174 9 analyses analysis NOUN bracis-19048 174 10 on on ADP bracis-19048 174 11 the the DET bracis-19048 174 12 meta meta ADJ bracis-19048 174 13 - - PUNCT bracis-19048 174 14 base base NOUN bracis-19048 174 15 involving involve VERB bracis-19048 174 16 the the DET bracis-19048 174 17 classifiers classifier NOUN bracis-19048 174 18 on on ADP bracis-19048 174 19 the the DET bracis-19048 174 20 selected select VERB bracis-19048 174 21 400 400 NUM bracis-19048 174 22 datasets dataset NOUN bracis-19048 174 23 . . PUNCT bracis-19048 175 1 fig fig NOUN bracis-19048 175 2 . . PUNCT bracis-19048 176 1 1 1 X bracis-19048 176 2 . . X bracis-19048 176 3 performance performance NOUN bracis-19048 176 4 of of ADP bracis-19048 176 5 classifiers classifier NOUN bracis-19048 176 6 over over ADP bracis-19048 176 7 the the DET bracis-19048 176 8 400 400 NUM bracis-19048 176 9 datasets dataset NOUN bracis-19048 176 10 . . PUNCT bracis-19048 177 1 full full ADJ bracis-19048 177 2 size size NOUN bracis-19048 177 3 image image NOUN bracis-19048 177 4 the the DET bracis-19048 177 5 distribution distribution NOUN bracis-19048 177 6 of of ADP bracis-19048 177 7 accuracy accuracy NOUN bracis-19048 177 8 values value NOUN bracis-19048 177 9 , , PUNCT bracis-19048 177 10 summarized summarize VERB bracis-19048 177 11 by by ADP bracis-19048 177 12 the the DET bracis-19048 177 13 boxplots boxplot NOUN bracis-19048 177 14 in in ADP bracis-19048 177 15 fig fig NOUN bracis-19048 177 16 . . PUNCT bracis-19048 177 17     SPACE bracis-19048 177 18 1a 1a NOUN bracis-19048 177 19 , , PUNCT bracis-19048 177 20 suggests suggest VERB bracis-19048 177 21 that that SCONJ bracis-19048 177 22 all all DET bracis-19048 177 23 approaches approach NOUN bracis-19048 177 24 have have VERB bracis-19048 177 25 generally generally ADV bracis-19048 177 26 similar similar ADJ bracis-19048 177 27 distributions distribution NOUN bracis-19048 177 28 of of ADP bracis-19048 177 29 performance performance NOUN bracis-19048 177 30 values value NOUN bracis-19048 177 31 across across ADP bracis-19048 177 32 the the DET bracis-19048 177 33 datasets dataset NOUN bracis-19048 177 34 employed employ VERB bracis-19048 177 35 , , PUNCT bracis-19048 177 36 with with ADP bracis-19048 177 37 reasonably reasonably ADV bracis-19048 177 38 high high ADJ bracis-19048 177 39 median median ADJ bracis-19048 177 40 accuracies accuracy NOUN bracis-19048 177 41 for for ADP bracis-19048 177 42 most most ADJ bracis-19048 177 43 problems problem NOUN bracis-19048 177 44 . . PUNCT bracis-19048 178 1 random random ADJ bracis-19048 178 2 forest forest NOUN bracis-19048 178 3 has have VERB bracis-19048 178 4 a a DET bracis-19048 178 5 slightly slightly ADV bracis-19048 178 6 higher high ADJ bracis-19048 178 7 median median ADJ bracis-19048 178 8 value value NOUN bracis-19048 178 9 than than ADP bracis-19048 178 10 the the DET bracis-19048 178 11 others other NOUN bracis-19048 178 12 , , PUNCT bracis-19048 178 13 while while SCONJ bracis-19048 178 14 knn knn PROPN bracis-19048 178 15 has have VERB bracis-19048 178 16 the the DET bracis-19048 178 17 lowest low ADJ bracis-19048 178 18 median median NOUN bracis-19048 178 19 and and CCONJ bracis-19048 178 20 the the DET bracis-19048 178 21 largest large ADJ bracis-19048 178 22 variability variability NOUN bracis-19048 178 23 in in ADP bracis-19048 178 24 performance performance NOUN bracis-19048 178 25 values value NOUN bracis-19048 178 26 . . PUNCT bracis-19048 179 1 figure figure NOUN bracis-19048 179 2     SPACE bracis-19048 179 3 1b 1b PROPN bracis-19048 179 4 shows show VERB bracis-19048 179 5 how how SCONJ bracis-19048 179 6 many many ADJ bracis-19048 179 7 times time NOUN bracis-19048 179 8 each each DET bracis-19048 179 9 algorithm algorithm NOUN bracis-19048 179 10 was be AUX bracis-19048 179 11 the the DET bracis-19048 179 12 “ " PUNCT bracis-19048 179 13 winner winner NOUN bracis-19048 179 14 ” " PUNCT bracis-19048 179 15 , , PUNCT bracis-19048 179 16 i.e. i.e. X bracis-19048 179 17 the the DET bracis-19048 179 18 number number NOUN bracis-19048 179 19 of of ADP bracis-19048 179 20 datasets dataset NOUN bracis-19048 179 21 for for ADP bracis-19048 179 22 which which PRON bracis-19048 179 23 each each PRON bracis-19048 179 24 presented present VERB bracis-19048 179 25 the the DET bracis-19048 179 26 best good ADJ bracis-19048 179 27 performance performance NOUN bracis-19048 179 28 . . PUNCT bracis-19048 180 1 these these DET bracis-19048 180 2 results result NOUN bracis-19048 180 3 imply imply VERB bracis-19048 180 4 that that SCONJ bracis-19048 180 5 the the DET bracis-19048 180 6 set set NOUN bracis-19048 180 7 of of ADP bracis-19048 180 8 choices choice NOUN bracis-19048 180 9 is be AUX bracis-19048 180 10 considered consider VERB bracis-19048 180 11 adequate adequate ADJ bracis-19048 180 12 for for ADP bracis-19048 180 13 mtl mtl PROPN bracis-19048 180 14 since since SCONJ bracis-19048 180 15 each each DET bracis-19048 180 16 one one NOUN bracis-19048 180 17 was be AUX bracis-19048 180 18 the the DET bracis-19048 180 19 best good ADJ bracis-19048 180 20 performing performing NOUN bracis-19048 180 21 in in ADP bracis-19048 180 22 a a DET bracis-19048 180 23 non non ADJ bracis-19048 180 24 - - ADJ bracis-19048 180 25 empty empty ADJ bracis-19048 180 26 subset subset NOUN bracis-19048 180 27 of of ADP bracis-19048 180 28 the the DET bracis-19048 180 29 dataset dataset NOUN bracis-19048 180 30 . . PUNCT bracis-19048 181 1 it it PRON bracis-19048 181 2 is be AUX bracis-19048 181 3 clear clear ADJ bracis-19048 181 4 from from ADP bracis-19048 181 5 the the DET bracis-19048 181 6 figures figure NOUN bracis-19048 181 7 that that PRON bracis-19048 181 8 rf rf VERB bracis-19048 181 9 was be AUX bracis-19048 181 10 most most ADV bracis-19048 181 11 frequently frequently ADV bracis-19048 181 12 the the DET bracis-19048 181 13 best good ADJ bracis-19048 181 14 classifier classifier NOUN bracis-19048 181 15 in in ADP bracis-19048 181 16 the the DET bracis-19048 181 17 dataset dataset NOUN bracis-19048 181 18 , , PUNCT bracis-19048 181 19 which which PRON bracis-19048 181 20 means mean VERB bracis-19048 181 21 it it PRON bracis-19048 181 22 will will AUX bracis-19048 181 23 be be AUX bracis-19048 181 24 the the DET bracis-19048 181 25 most most ADV bracis-19048 181 26 probable probable ADJ bracis-19048 181 27 choice choice NOUN bracis-19048 181 28 for for ADP bracis-19048 181 29 the the DET bracis-19048 181 30 random random ADJ bracis-19048 181 31 baseline baseline NOUN bracis-19048 181 32 recommender recommender NOUN bracis-19048 181 33 system system NOUN bracis-19048 181 34 and and CCONJ bracis-19048 181 35 the the DET bracis-19048 181 36 fixed fix VERB bracis-19048 181 37 choice choice NOUN bracis-19048 181 38 for for ADP bracis-19048 181 39 the the DET bracis-19048 181 40 default default NOUN bracis-19048 181 41 baseline baseline NOUN bracis-19048 181 42 one one NUM bracis-19048 181 43 . . PUNCT bracis-19048 182 1 to to PART bracis-19048 182 2 investigate investigate VERB bracis-19048 182 3 their their PRON bracis-19048 182 4 quality quality NOUN bracis-19048 182 5 on on ADP bracis-19048 182 6 the the DET bracis-19048 182 7 task task NOUN bracis-19048 182 8 of of ADP bracis-19048 182 9 learning learn VERB bracis-19048 182 10 to to PART bracis-19048 182 11 recommend recommend VERB bracis-19048 182 12 a a DET bracis-19048 182 13 classifier classifier NOUN bracis-19048 182 14 , , PUNCT bracis-19048 182 15 the the DET bracis-19048 182 16 selected select VERB bracis-19048 182 17 regressors regressor NOUN bracis-19048 182 18 ( ( PUNCT bracis-19048 182 19 mlpr mlpr PROPN bracis-19048 182 20 , , PUNCT bracis-19048 182 21 cart cart NOUN bracis-19048 182 22 , , PUNCT bracis-19048 182 23 dwnn dwnn NOUN bracis-19048 182 24 , , PUNCT bracis-19048 182 25 svr svr PROPN bracis-19048 182 26 , , PUNCT bracis-19048 182 27 and and CCONJ bracis-19048 182 28 rfr rfr PROPN bracis-19048 182 29 ) ) PUNCT bracis-19048 182 30 and and CCONJ bracis-19048 182 31 the the DET bracis-19048 182 32 two two NUM bracis-19048 182 33 baselines baseline NOUN bracis-19048 182 34 regressors regressor NOUN bracis-19048 182 35 ( ( PUNCT bracis-19048 182 36 rd rd NOUN bracis-19048 182 37 and and CCONJ bracis-19048 182 38 df df PROPN bracis-19048 182 39 ) ) PUNCT bracis-19048 182 40 were be AUX bracis-19048 182 41 applied apply VERB bracis-19048 182 42 to to ADP bracis-19048 182 43 the the DET bracis-19048 182 44 meta meta ADV bracis-19048 182 45 - - PUNCT bracis-19048 182 46 dataset dataset ADJ bracis-19048 182 47 considering consider VERB bracis-19048 182 48 two two NUM bracis-19048 182 49 sets set NOUN bracis-19048 182 50 of of ADP bracis-19048 182 51 meta meta NOUN bracis-19048 182 52 - - PUNCT bracis-19048 182 53 features feature NOUN bracis-19048 182 54 : : PUNCT bracis-19048 182 55 the the DET bracis-19048 182 56 112 112 NUM bracis-19048 182 57 standard standard ADJ bracis-19048 182 58 meta meta NOUN bracis-19048 182 59 - - PUNCT bracis-19048 182 60 features feature NOUN bracis-19048 182 61 ( ( PUNCT bracis-19048 182 62 described describe VERB bracis-19048 182 63 in in ADP bracis-19048 182 64 sect sect NOUN bracis-19048 182 65 . . PUNCT bracis-19048 182 66     SPACE bracis-19048 182 67 2.1 2.1 NUM bracis-19048 182 68 ) ) PUNCT bracis-19048 182 69 and and CCONJ bracis-19048 182 70 the the DET bracis-19048 182 71 full full ADJ bracis-19048 182 72 set set NOUN bracis-19048 182 73 of of ADP bracis-19048 182 74 130 130 NUM bracis-19048 182 75 meta meta NOUN bracis-19048 182 76 - - PUNCT bracis-19048 182 77 features feature NOUN bracis-19048 182 78 ( ( PUNCT bracis-19048 182 79 incorporating incorporate VERB bracis-19048 182 80 the the DET bracis-19048 182 81 18 18 NUM bracis-19048 182 82 clustering clustering ADJ bracis-19048 182 83 measures measure NOUN bracis-19048 182 84 from from ADP bracis-19048 182 85 sect sect NOUN bracis-19048 182 86 . . PUNCT bracis-19048 182 87     SPACE bracis-19048 182 88 2.2 2.2 NUM bracis-19048 182 89 ) ) PUNCT bracis-19048 182 90 . . PUNCT bracis-19048 183 1 figure figure NOUN bracis-19048 183 2     SPACE bracis-19048 183 3 2 2 NUM bracis-19048 183 4 shows show VERB bracis-19048 183 5 the the DET bracis-19048 183 6 average average NOUN bracis-19048 183 7 normalized normalize VERB bracis-19048 183 8 mean mean VERB bracis-19048 183 9 squared square VERB bracis-19048 183 10 error error NOUN bracis-19048 183 11 ( ( PUNCT bracis-19048 183 12 nmse nmse NOUN bracis-19048 183 13 ) ) PUNCT bracis-19048 183 14 of of ADP bracis-19048 183 15 each each DET bracis-19048 183 16 meta meta NOUN bracis-19048 183 17 - - PUNCT bracis-19048 183 18 regressor regressor NOUN bracis-19048 183 19 on on ADP bracis-19048 183 20 predicting predict VERB bracis-19048 183 21 the the DET bracis-19048 183 22 expected expect VERB bracis-19048 183 23 accuracy accuracy NOUN bracis-19048 183 24 of of ADP bracis-19048 183 25 the the DET bracis-19048 183 26 classifiers classifier NOUN bracis-19048 183 27 for for ADP bracis-19048 183 28 each each DET bracis-19048 183 29 set set NOUN bracis-19048 183 30 of of ADP bracis-19048 183 31 meta meta NOUN bracis-19048 183 32 - - PUNCT bracis-19048 183 33 features feature NOUN bracis-19048 183 34 , , PUNCT bracis-19048 183 35 estimated estimate VERB bracis-19048 183 36 by by ADP bracis-19048 183 37 10 10 NUM bracis-19048 183 38 - - ADJ bracis-19048 183 39 fold fold ADJ bracis-19048 183 40 cross cross NOUN bracis-19048 183 41 - - NOUN bracis-19048 183 42 validation validation ADJ bracis-19048 183 43 . . PUNCT bracis-19048 184 1 fig fig NOUN bracis-19048 184 2 . . PUNCT bracis-19048 185 1 2 2 X bracis-19048 185 2 . . X bracis-19048 185 3 the the DET bracis-19048 185 4 log log NOUN bracis-19048 185 5 - - PUNCT bracis-19048 185 6 scaled scale VERB bracis-19048 185 7 nmse nmse NOUN bracis-19048 185 8 for for ADP bracis-19048 185 9 each each DET bracis-19048 185 10 combination combination NOUN bracis-19048 185 11 of of ADP bracis-19048 185 12 meta meta ADJ bracis-19048 185 13 - - PUNCT bracis-19048 185 14 feature feature NOUN bracis-19048 185 15 set set NOUN bracis-19048 185 16 , , PUNCT bracis-19048 185 17 classification classification NOUN bracis-19048 185 18 and and CCONJ bracis-19048 185 19 regression regression NOUN bracis-19048 185 20 models model NOUN bracis-19048 185 21 . . PUNCT bracis-19048 186 1 full full ADJ bracis-19048 186 2 size size NOUN bracis-19048 186 3 image image NOUN bracis-19048 186 4 the the DET bracis-19048 186 5 boxplots boxplot NOUN bracis-19048 186 6 indicate indicate VERB bracis-19048 186 7 that that SCONJ bracis-19048 186 8 all all DET bracis-19048 186 9 meta meta NOUN bracis-19048 186 10 - - PUNCT bracis-19048 186 11 regressors regressor NOUN bracis-19048 186 12 provided provide VERB bracis-19048 186 13 better well ADJ bracis-19048 186 14 predictions prediction NOUN bracis-19048 186 15 than than ADP bracis-19048 186 16 the the DET bracis-19048 186 17 baseline baseline NOUN bracis-19048 186 18 approaches approach NOUN bracis-19048 186 19 ( ( PUNCT bracis-19048 186 20 with with ADP bracis-19048 186 21 the the DET bracis-19048 186 22 exception exception NOUN bracis-19048 186 23 of of ADP bracis-19048 186 24 a a DET bracis-19048 186 25 single single ADJ bracis-19048 186 26 observation observation NOUN bracis-19048 186 27 for for ADP bracis-19048 186 28 the the DET bracis-19048 186 29 mlpr mlpr PROPN bracis-19048 186 30 regressor regressor PROPN bracis-19048 186 31 ) ) PUNCT bracis-19048 186 32 . . PUNCT bracis-19048 187 1 rfr rfr PROPN bracis-19048 187 2 seems seem VERB bracis-19048 187 3 to to PART bracis-19048 187 4 have have VERB bracis-19048 187 5 a a DET bracis-19048 187 6 slight slight ADJ bracis-19048 187 7 advantage advantage NOUN bracis-19048 187 8 when when SCONJ bracis-19048 187 9 compared compare VERB bracis-19048 187 10 to to ADP bracis-19048 187 11 the the DET bracis-19048 187 12 other other ADJ bracis-19048 187 13 regressors regressor NOUN bracis-19048 187 14 , , PUNCT bracis-19048 187 15 for for ADP bracis-19048 187 16 both both DET bracis-19048 187 17 sets set NOUN bracis-19048 187 18 of of ADP bracis-19048 187 19 features feature NOUN bracis-19048 187 20 . . PUNCT bracis-19048 188 1 mlpr mlpr PROPN bracis-19048 188 2 , , PUNCT bracis-19048 188 3 cart cart NOUN bracis-19048 188 4 , , PUNCT bracis-19048 188 5 dwnn dwnn NOUN bracis-19048 188 6 and and CCONJ bracis-19048 188 7 svr svr PROPN bracis-19048 188 8 show show VERB bracis-19048 188 9 a a DET bracis-19048 188 10 similar similar ADJ bracis-19048 188 11 performance performance NOUN bracis-19048 188 12 distribution distribution NOUN bracis-19048 188 13 in in ADP bracis-19048 188 14 most most ADJ bracis-19048 188 15 cases case NOUN bracis-19048 188 16 , , PUNCT bracis-19048 188 17 with with ADP bracis-19048 188 18 mlpr mlpr PROPN bracis-19048 188 19 having have VERB bracis-19048 188 20 a a DET bracis-19048 188 21 higher high ADJ bracis-19048 188 22 interquartile interquartile NOUN bracis-19048 188 23 range range NOUN bracis-19048 188 24 and and CCONJ bracis-19048 188 25 a a DET bracis-19048 188 26 few few ADJ bracis-19048 188 27 outliers outlier NOUN bracis-19048 188 28 . . PUNCT bracis-19048 189 1 although although SCONJ bracis-19048 189 2 visual visual ADJ bracis-19048 189 3 inspection inspection NOUN bracis-19048 189 4 does do AUX bracis-19048 189 5 not not PART bracis-19048 189 6 necessarily necessarily ADV bracis-19048 189 7 provide provide VERB bracis-19048 189 8 a a DET bracis-19048 189 9 clear clear ADJ bracis-19048 189 10 winner winner NOUN bracis-19048 189 11 between between ADP bracis-19048 189 12 the the DET bracis-19048 189 13 two two NUM bracis-19048 189 14 sets set NOUN bracis-19048 189 15 of of ADP bracis-19048 189 16 meta meta NOUN bracis-19048 189 17 - - PUNCT bracis-19048 189 18 features feature NOUN bracis-19048 189 19 , , PUNCT bracis-19048 189 20 it it PRON bracis-19048 189 21 is be AUX bracis-19048 189 22 sufficient sufficient ADJ bracis-19048 189 23 to to PART bracis-19048 189 24 indicate indicate VERB bracis-19048 189 25 that that PRON bracis-19048 189 26 : : PUNCT bracis-19048 189 27 ( ( PUNCT bracis-19048 189 28 i i NOUN bracis-19048 189 29 ) ) PUNCT bracis-19048 189 30 they they PRON bracis-19048 189 31 are be AUX bracis-19048 189 32 clearly clearly ADV bracis-19048 189 33 less less ADJ bracis-19048 189 34 error error NOUN bracis-19048 189 35 - - PUNCT bracis-19048 189 36 prone prone ADJ bracis-19048 189 37 than than ADP bracis-19048 189 38 the the DET bracis-19048 189 39 baselines baseline NOUN bracis-19048 189 40 , , PUNCT bracis-19048 189 41 and and CCONJ bracis-19048 189 42 ( ( PUNCT bracis-19048 189 43 ii ii NOUN bracis-19048 189 44 ) ) PUNCT bracis-19048 189 45 their their PRON bracis-19048 189 46 nmse nmse ADJ bracis-19048 189 47 values value NOUN bracis-19048 189 48 show show VERB bracis-19048 189 49 a a DET bracis-19048 189 50 relatively relatively ADV bracis-19048 189 51 low low ADJ bracis-19048 189 52 variance variance NOUN bracis-19048 189 53 . . PUNCT bracis-19048 190 1 more more ADV bracis-19048 190 2 objectively objectively ADV bracis-19048 190 3 , , PUNCT bracis-19048 190 4 an an DET bracis-19048 190 5 anova anova PROPN bracis-19048 190 6 model model NOUN bracis-19048 190 7 followed follow VERB bracis-19048 190 8 by by ADP bracis-19048 190 9 dunnett dunnett PROPN bracis-19048 190 10 pairwise pairwise PROPN bracis-19048 190 11 comparisons comparison NOUN bracis-19048 190 12 indicates indicate VERB bracis-19048 190 13 a a DET bracis-19048 190 14 statistically statistically ADV bracis-19048 190 15 significant significant ADJ bracis-19048 190 16 difference difference NOUN bracis-19048 190 17 at at ADP bracis-19048 190 18 the the DET bracis-19048 190 19 \(95\%\ \(95\%\ NOUN bracis-19048 190 20 ) ) PUNCT bracis-19048 190 21 confidence confidence NOUN bracis-19048 190 22 level level NOUN bracis-19048 190 23 between between ADP bracis-19048 190 24 the the DET bracis-19048 190 25 mtl mtl PROPN bracis-19048 190 26 algorithms algorithms NOUN bracis-19048 190 27 ( ( PUNCT bracis-19048 190 28 in in ADP bracis-19048 190 29 aggregate aggregate NOUN bracis-19048 190 30 ) ) PUNCT bracis-19048 190 31 against against ADP bracis-19048 190 32 the the DET bracis-19048 190 33 two two NUM bracis-19048 190 34 baselines baseline NOUN bracis-19048 190 35 ( ( PUNCT bracis-19048 190 36 \(p \(p X bracis-19048 190 37 < < X bracis-19048 190 38 10^{-10}\ 10^{-10}\ NUM bracis-19048 190 39 ) ) PUNCT bracis-19048 190 40 for for ADP bracis-19048 190 41 both both CCONJ bracis-19048 190 42 the the DET bracis-19048 190 43 mtl\(\times mtl\(\time NOUN bracis-19048 190 44 \)df \)df NOUN bracis-19048 190 45 and and CCONJ bracis-19048 190 46 mtl\(\times mtl\(\time NOUN bracis-19048 190 47 \)rd \)rd ADJ bracis-19048 190 48 comparisons comparison NOUN bracis-19048 190 49 ) ) PUNCT bracis-19048 190 50 . . PUNCT bracis-19048 191 1 the the DET bracis-19048 191 2 effects effect NOUN bracis-19048 191 3 of of ADP bracis-19048 191 4 different different ADJ bracis-19048 191 5 meta meta ADJ bracis-19048 191 6 - - PUNCT bracis-19048 191 7 feature feature NOUN bracis-19048 191 8 sets set NOUN bracis-19048 191 9 , , PUNCT bracis-19048 191 10 classifier classifier NOUN bracis-19048 191 11 methods method NOUN bracis-19048 191 12 and and CCONJ bracis-19048 191 13 distinct distinct ADJ bracis-19048 191 14 replicates replicate NOUN bracis-19048 191 15 were be AUX bracis-19048 191 16 removed remove VERB bracis-19048 191 17 by by ADP bracis-19048 191 18 blocking block VERB bracis-19048 191 19     SPACE bracis-19048 191 20 [ [ X bracis-19048 191 21 23 23 NUM bracis-19048 191 22 ] ] PUNCT bracis-19048 191 23 . . PUNCT bracis-19048 192 1 a a DET bracis-19048 192 2 second second ADJ bracis-19048 192 3 analysis analysis NOUN bracis-19048 192 4 was be AUX bracis-19048 192 5 conducted conduct VERB bracis-19048 192 6 to to PART bracis-19048 192 7 investigate investigate VERB bracis-19048 192 8 the the DET bracis-19048 192 9 effect effect NOUN bracis-19048 192 10 of of ADP bracis-19048 192 11 adding add VERB bracis-19048 192 12 the the DET bracis-19048 192 13 clustering clustering ADJ bracis-19048 192 14 meta meta NOUN bracis-19048 192 15 - - PUNCT bracis-19048 192 16 features feature NOUN bracis-19048 192 17 on on ADP bracis-19048 192 18 the the DET bracis-19048 192 19 predictive predictive ADJ bracis-19048 192 20 ability ability NOUN bracis-19048 192 21 of of ADP bracis-19048 192 22 the the DET bracis-19048 192 23 meta meta NOUN bracis-19048 192 24 - - PUNCT bracis-19048 192 25 regressors regressor NOUN bracis-19048 192 26 . . PUNCT bracis-19048 193 1 this this DET bracis-19048 193 2 analysis analysis NOUN bracis-19048 193 3 was be AUX bracis-19048 193 4 performed perform VERB bracis-19048 193 5 by by ADP bracis-19048 193 6 removing remove VERB bracis-19048 193 7 the the DET bracis-19048 193 8 two two NUM bracis-19048 193 9 baseline baseline ADJ bracis-19048 193 10 methods method NOUN bracis-19048 193 11 and and CCONJ bracis-19048 193 12 fitting fit VERB bracis-19048 193 13 a a DET bracis-19048 193 14 blocked block VERB bracis-19048 193 15 anova anova PROPN bracis-19048 193 16 model model PROPN bracis-19048 193 17     SPACE bracis-19048 194 1 [ [ X bracis-19048 194 2 23 23 NUM bracis-19048 194 3 ] ] PUNCT bracis-19048 194 4 on on ADP bracis-19048 194 5 rank rank NOUN bracis-19048 194 6 - - PUNCT bracis-19048 194 7 transformed transform VERB bracis-19048 194 8 data datum NOUN bracis-19048 194 9 ( ( PUNCT bracis-19048 194 10 required require VERB bracis-19048 194 11 in in ADP bracis-19048 194 12 this this DET bracis-19048 194 13 case case NOUN bracis-19048 194 14 to to PART bracis-19048 194 15 meet meet VERB bracis-19048 194 16 the the DET bracis-19048 194 17 anova anova PROPN bracis-19048 194 18 assumptions assumption NOUN bracis-19048 194 19 ) ) PUNCT bracis-19048 194 20 , , PUNCT bracis-19048 194 21 with with ADP bracis-19048 194 22 the the DET bracis-19048 194 23 meta meta ADJ bracis-19048 194 24 - - PUNCT bracis-19048 194 25 feature feature NOUN bracis-19048 194 26 sets set NOUN bracis-19048 194 27 as as ADP bracis-19048 194 28 the the DET bracis-19048 194 29 experimental experimental ADJ bracis-19048 194 30 factor factor NOUN bracis-19048 194 31 and and CCONJ bracis-19048 194 32 the the DET bracis-19048 194 33 classifier classifier NOUN bracis-19048 194 34 methods method NOUN bracis-19048 194 35 , , PUNCT bracis-19048 194 36 meta meta NOUN bracis-19048 194 37 - - PUNCT bracis-19048 194 38 regressors regressor NOUN bracis-19048 194 39 , , PUNCT bracis-19048 194 40 and and CCONJ bracis-19048 194 41 replicates replicate NOUN bracis-19048 194 42 as as ADP bracis-19048 194 43 blocking block VERB bracis-19048 194 44 factors factor NOUN bracis-19048 194 45 . . PUNCT bracis-19048 195 1 this this DET bracis-19048 195 2 test test NOUN bracis-19048 195 3 reveal reveal VERB bracis-19048 195 4 a a DET bracis-19048 195 5 statistically statistically ADV bracis-19048 195 6 significant significant ADJ bracis-19048 195 7 positive positive ADJ bracis-19048 195 8 effect effect NOUN bracis-19048 195 9 of of ADP bracis-19048 195 10 adding add VERB bracis-19048 195 11 the the DET bracis-19048 195 12 clustering clustering ADJ bracis-19048 195 13 ones one NOUN bracis-19048 195 14 on on ADP bracis-19048 195 15 the the DET bracis-19048 195 16 performance performance NOUN bracis-19048 195 17 of of ADP bracis-19048 195 18 the the DET bracis-19048 195 19 best good ADJ bracis-19048 195 20 regressor regressor NOUN bracis-19048 195 21 , , PUNCT bracis-19048 195 22 rfr rfr PROPN bracis-19048 195 23 , , PUNCT bracis-19048 195 24 across across ADP bracis-19048 195 25 all all DET bracis-19048 195 26 classifiers classifier NOUN bracis-19048 195 27 ( ( PUNCT bracis-19048 195 28 paired pair VERB bracis-19048 195 29 t t NOUN bracis-19048 195 30 - - PUNCT bracis-19048 195 31 test test NOUN bracis-19048 195 32 for for ADP bracis-19048 195 33 standard\(\times standard\(\time NOUN bracis-19048 195 34 \)full \)full VERB bracis-19048 195 35 meta meta ADJ bracis-19048 195 36 - - PUNCT bracis-19048 195 37 feature feature NOUN bracis-19048 195 38 sets set NOUN bracis-19048 195 39 , , PUNCT bracis-19048 195 40 \(p \(p PROPN bracis-19048 195 41 = = SYM bracis-19048 195 42 0.036\ 0.036\ PROPN bracis-19048 195 43 ) ) PUNCT bracis-19048 195 44 ) ) PUNCT bracis-19048 195 45 . . PUNCT bracis-19048 196 1 figure figure NOUN bracis-19048 196 2     SPACE bracis-19048 196 3 3 3 NUM bracis-19048 196 4 shows show VERB bracis-19048 196 5 the the DET bracis-19048 196 6 results result NOUN bracis-19048 196 7 of of ADP bracis-19048 196 8 a a DET bracis-19048 196 9 direct direct ADJ bracis-19048 196 10 comparison comparison NOUN bracis-19048 196 11 between between ADP bracis-19048 196 12 the the DET bracis-19048 196 13 recommended recommend VERB bracis-19048 196 14 classifier classifier NOUN bracis-19048 196 15 . . PUNCT bracis-19048 197 1 the the DET bracis-19048 197 2 x x ADJ bracis-19048 197 3 - - ADJ bracis-19048 197 4 axis axis ADJ bracis-19048 197 5 lists list NOUN bracis-19048 197 6 the the DET bracis-19048 197 7 meta meta NOUN bracis-19048 197 8 - - PUNCT bracis-19048 197 9 regressors regressor NOUN bracis-19048 197 10 and and CCONJ bracis-19048 197 11 the the DET bracis-19048 197 12 y y ADJ bracis-19048 197 13 - - PUNCT bracis-19048 197 14 axis axis NOUN bracis-19048 197 15 shows show VERB bracis-19048 197 16 the the DET bracis-19048 197 17 percent percent NOUN bracis-19048 197 18 differences difference NOUN bracis-19048 197 19 in in ADP bracis-19048 197 20 accuracy accuracy NOUN bracis-19048 197 21 , , PUNCT bracis-19048 197 22 averaged average VERB bracis-19048 197 23 over over ADP bracis-19048 197 24 all all DET bracis-19048 197 25 datasets dataset NOUN bracis-19048 197 26 . . PUNCT bracis-19048 198 1 as as SCONJ bracis-19048 198 2 seen see VERB bracis-19048 198 3 in in ADP bracis-19048 198 4 the the DET bracis-19048 198 5 analysis analysis NOUN bracis-19048 198 6 of of ADP bracis-19048 198 7 fig fig NOUN bracis-19048 198 8 . . PUNCT bracis-19048 198 9     SPACE bracis-19048 199 1 2 2 NUM bracis-19048 199 2 there there PRON bracis-19048 199 3 is be VERB bracis-19048 199 4 a a DET bracis-19048 199 5 noticeable noticeable ADJ bracis-19048 199 6 advantage advantage NOUN bracis-19048 199 7 of of ADP bracis-19048 199 8 using use VERB bracis-19048 199 9 the the DET bracis-19048 199 10 mtl mtl PROPN bracis-19048 199 11 approaches approach NOUN bracis-19048 199 12 over over ADP bracis-19048 199 13 the the DET bracis-19048 199 14 random random ADJ bracis-19048 199 15 baseline baseline NOUN bracis-19048 199 16 ( ( PUNCT bracis-19048 199 17 rd rd NOUN bracis-19048 199 18 ) ) PUNCT bracis-19048 199 19 , , PUNCT bracis-19048 199 20 as as ADV bracis-19048 199 21 well well ADV bracis-19048 199 22 as as ADP bracis-19048 199 23 against against ADP bracis-19048 199 24 the the DET bracis-19048 199 25 fixed fix VERB bracis-19048 199 26 choice choice NOUN bracis-19048 199 27 of of ADP bracis-19048 199 28 rf rf NOUN bracis-19048 199 29 , , PUNCT bracis-19048 199 30 which which PRON bracis-19048 199 31 is be AUX bracis-19048 199 32 best good ADJ bracis-19048 199 33 overall overall ADJ bracis-19048 199 34 classifier classifier NOUN bracis-19048 199 35 ( ( PUNCT bracis-19048 199 36 df df PROPN bracis-19048 199 37 ) ) PUNCT bracis-19048 199 38 , , PUNCT bracis-19048 199 39 in in ADP bracis-19048 199 40 the the DET bracis-19048 199 41 case case NOUN bracis-19048 199 42 of of ADP bracis-19048 199 43 the the DET bracis-19048 199 44 three three NUM bracis-19048 199 45 most most ADV bracis-19048 199 46 successful successful ADJ bracis-19048 199 47 regressors regressor NOUN bracis-19048 199 48 . . PUNCT bracis-19048 200 1 the the DET bracis-19048 200 2 magnitude magnitude NOUN bracis-19048 200 3 of of ADP bracis-19048 200 4 the the DET bracis-19048 200 5 gains gain NOUN bracis-19048 200 6 is be AUX bracis-19048 200 7 also also ADV bracis-19048 200 8 substantial substantial ADJ bracis-19048 200 9 , , PUNCT bracis-19048 200 10 with with ADP bracis-19048 200 11 40–\(50\%\ 40–\(50\%\ NOUN bracis-19048 200 12 ) ) PUNCT bracis-19048 200 13 gains gain NOUN bracis-19048 200 14 over over ADP bracis-19048 200 15 rd rd PROPN bracis-19048 200 16 and and CCONJ bracis-19048 200 17 10–\(20\%\ 10–\(20\%\ NUM bracis-19048 200 18 ) ) PUNCT bracis-19048 200 19 gains gain NOUN bracis-19048 200 20 over over ADP bracis-19048 200 21 df df NOUN bracis-19048 200 22 in in ADP bracis-19048 200 23 the the DET bracis-19048 200 24 case case NOUN bracis-19048 200 25 of of ADP bracis-19048 200 26 dwnn dwnn NOUN bracis-19048 200 27 , , PUNCT bracis-19048 200 28 rfr rfr PROPN bracis-19048 200 29 and and CCONJ bracis-19048 200 30 svr svr PROPN bracis-19048 200 31 . . PROPN bracis-19048 200 32 fig fig PROPN bracis-19048 200 33 . . PUNCT bracis-19048 201 1 3 3 X bracis-19048 201 2 . . X bracis-19048 201 3 differences difference NOUN bracis-19048 201 4 between between ADP bracis-19048 201 5 base base NOUN bracis-19048 201 6 - - PUNCT bracis-19048 201 7 classifier classifier NOUN bracis-19048 201 8 accuracies accuracy NOUN bracis-19048 201 9 over over ADP bracis-19048 201 10 baselines baseline NOUN bracis-19048 201 11 . . PUNCT bracis-19048 202 1 vertical vertical ADJ bracis-19048 202 2 lines line NOUN bracis-19048 202 3 represent represent VERB bracis-19048 202 4 standard standard ADJ bracis-19048 202 5 errors error NOUN bracis-19048 202 6 . . PUNCT bracis-19048 203 1 full full ADJ bracis-19048 203 2 size size NOUN bracis-19048 203 3 image image VERB bracis-19048 203 4 the the DET bracis-19048 203 5 impact impact NOUN bracis-19048 203 6 of of ADP bracis-19048 203 7 the the DET bracis-19048 203 8 meta meta NOUN bracis-19048 203 9 - - PUNCT bracis-19048 203 10 features feature NOUN bracis-19048 203 11 can can AUX bracis-19048 203 12 be be AUX bracis-19048 203 13 further far ADV bracis-19048 203 14 investigated investigate VERB bracis-19048 203 15 via via ADP bracis-19048 203 16 the the DET bracis-19048 203 17 analysis analysis NOUN bracis-19048 203 18 of of ADP bracis-19048 203 19 the the DET bracis-19048 203 20 rfr rfr PROPN bracis-19048 203 21 model model NOUN bracis-19048 203 22 . . PUNCT bracis-19048 204 1 figure figure VERB bracis-19048 204 2     SPACE bracis-19048 204 3 4 4 NUM bracis-19048 204 4 shows show VERB bracis-19048 204 5 the the DET bracis-19048 204 6 most most ADV bracis-19048 204 7 relevant relevant ADJ bracis-19048 204 8 meta meta NOUN bracis-19048 204 9 - - PUNCT bracis-19048 204 10 features feature NOUN bracis-19048 204 11 , , PUNCT bracis-19048 204 12 according accord VERB bracis-19048 204 13 to to ADP bracis-19048 204 14 their their PRON bracis-19048 204 15 individual individual ADJ bracis-19048 204 16 contribution contribution NOUN bracis-19048 204 17 to to ADP bracis-19048 204 18 the the DET bracis-19048 204 19 reduction reduction NOUN bracis-19048 204 20 of of ADP bracis-19048 204 21 the the DET bracis-19048 204 22 mean mean ADJ bracis-19048 204 23 squared square VERB bracis-19048 204 24 error error NOUN bracis-19048 204 25 , , PUNCT bracis-19048 204 26 mse mse PROPN bracis-19048 204 27 ( ( PUNCT bracis-19048 204 28 measured measure VERB bracis-19048 204 29 as as ADP bracis-19048 204 30 the the DET bracis-19048 204 31 increase increase NOUN bracis-19048 204 32 in in ADP bracis-19048 204 33 mse mse NOUN bracis-19048 204 34 when when SCONJ bracis-19048 204 35 that that SCONJ bracis-19048 204 36 meta meta ADJ bracis-19048 204 37 - - PUNCT bracis-19048 204 38 feature feature NOUN bracis-19048 204 39 was be AUX bracis-19048 204 40 omitted omit VERB bracis-19048 204 41 ) ) PUNCT bracis-19048 204 42 , , PUNCT bracis-19048 204 43 as as ADV bracis-19048 204 44 well well ADV bracis-19048 204 45 as as ADP bracis-19048 204 46 their their PRON bracis-19048 204 47 corresponding correspond VERB bracis-19048 204 48 groups group NOUN bracis-19048 204 49 . . PUNCT bracis-19048 205 1 the the DET bracis-19048 205 2 x x ADJ bracis-19048 205 3 - - ADJ bracis-19048 205 4 axis axis ADJ bracis-19048 205 5 lists list NOUN bracis-19048 205 6 the the DET bracis-19048 205 7 meta meta NOUN bracis-19048 205 8 - - PUNCT bracis-19048 205 9 features feature NOUN bracis-19048 205 10 in in ADP bracis-19048 205 11 decreasing decrease VERB bracis-19048 205 12 order order NOUN bracis-19048 205 13 of of ADP bracis-19048 205 14 relevance relevance NOUN bracis-19048 205 15 , , PUNCT bracis-19048 205 16 with with ADP bracis-19048 205 17 the the DET bracis-19048 205 18 mean mean ADJ bracis-19048 205 19 change change NOUN bracis-19048 205 20 in in ADP bracis-19048 205 21 mse mse NOUN bracis-19048 205 22 shown show VERB bracis-19048 205 23 in in ADP bracis-19048 205 24 the the DET bracis-19048 205 25 y y NOUN bracis-19048 205 26 - - PUNCT bracis-19048 205 27 axis axis NOUN bracis-19048 205 28 . . PUNCT bracis-19048 206 1 vertical vertical ADJ bracis-19048 206 2 bars bar NOUN bracis-19048 206 3 indicate indicate VERB bracis-19048 206 4 standard standard ADJ bracis-19048 206 5 errors error NOUN bracis-19048 206 6 of of ADP bracis-19048 206 7 estimation estimation NOUN bracis-19048 206 8 . . PUNCT bracis-19048 207 1 fig fig NOUN bracis-19048 207 2 . . PUNCT bracis-19048 208 1 4 4 X bracis-19048 208 2 . . X bracis-19048 208 3 top top ADV bracis-19048 208 4 - - PUNCT bracis-19048 208 5 ranked rank VERB bracis-19048 208 6 meta meta NOUN bracis-19048 208 7 - - PUNCT bracis-19048 208 8 features feature NOUN bracis-19048 208 9 selected select VERB bracis-19048 208 10 by by ADP bracis-19048 208 11 the the DET bracis-19048 208 12 rfrs rfrs NOUN bracis-19048 208 13 in in ADP bracis-19048 208 14 log log NOUN bracis-19048 208 15 - - PUNCT bracis-19048 208 16 scale scale NOUN bracis-19048 208 17 , , PUNCT bracis-19048 208 18 based base VERB bracis-19048 208 19 on on ADP bracis-19048 208 20 the the DET bracis-19048 208 21 average average ADJ bracis-19048 208 22 increase increase NOUN bracis-19048 208 23 in in ADP bracis-19048 208 24 mse mse NOUN bracis-19048 208 25 when when SCONJ bracis-19048 208 26 the the DET bracis-19048 208 27 meta meta ADJ bracis-19048 208 28 - - PUNCT bracis-19048 208 29 feature feature NOUN bracis-19048 208 30 is be AUX bracis-19048 208 31 omitted omit VERB bracis-19048 208 32 . . PUNCT bracis-19048 209 1 full full ADJ bracis-19048 209 2 size size NOUN bracis-19048 209 3 image image NOUN bracis-19048 209 4 the the DET bracis-19048 209 5 standard standard ADJ bracis-19048 209 6 meta meta NOUN bracis-19048 209 7 - - PUNCT bracis-19048 209 8 features feature NOUN bracis-19048 209 9 are be AUX bracis-19048 209 10 more more ADV bracis-19048 209 11 strongly strongly ADV bracis-19048 209 12 represented represent VERB bracis-19048 209 13 , , PUNCT bracis-19048 209 14 as as SCONJ bracis-19048 209 15 expected expect VERB bracis-19048 209 16 due due ADP bracis-19048 209 17 to to ADP bracis-19048 209 18 their their PRON bracis-19048 209 19 informative informative ADJ bracis-19048 209 20 value value NOUN bracis-19048 209 21     SPACE bracis-19048 210 1 [ [ X bracis-19048 210 2 31 31 NUM bracis-19048 210 3 ] ] PUNCT bracis-19048 210 4 . . PUNCT bracis-19048 211 1 some some DET bracis-19048 211 2 meta meta ADJ bracis-19048 211 3 - - PUNCT bracis-19048 211 4 features feature NOUN bracis-19048 211 5 appear appear VERB bracis-19048 211 6 more more ADV bracis-19048 211 7 than than ADP bracis-19048 211 8 once once ADV bracis-19048 211 9 due due ADJ bracis-19048 211 10 to to ADP bracis-19048 211 11 distinct distinct ADJ bracis-19048 211 12 summarization summarization NOUN bracis-19048 211 13 functions function NOUN bracis-19048 211 14 being be AUX bracis-19048 211 15 used use VERB bracis-19048 211 16 . . PUNCT bracis-19048 212 1 landmarking landmarking NOUN bracis-19048 212 2 measures measure NOUN bracis-19048 212 3 , , PUNCT bracis-19048 212 4 which which PRON bracis-19048 212 5 mainly mainly ADV bracis-19048 212 6 relate relate VERB bracis-19048 212 7 to to ADP bracis-19048 212 8 the the DET bracis-19048 212 9 performance performance NOUN bracis-19048 212 10 of of ADP bracis-19048 212 11 simple simple ADJ bracis-19048 212 12 meta meta ADJ bracis-19048 212 13 - - PUNCT bracis-19048 212 14 models model NOUN bracis-19048 212 15 induced induce VERB bracis-19048 212 16 by by ADP bracis-19048 212 17 the the DET bracis-19048 212 18 knn knn PROPN bracis-19048 212 19 , , PUNCT bracis-19048 212 20 naïve naïve ADJ bracis-19048 212 21 bayes baye NOUN bracis-19048 212 22 and and CCONJ bracis-19048 212 23 simple simple ADJ bracis-19048 212 24 node node NOUN bracis-19048 212 25 dt dt NOUN bracis-19048 212 26 algorithms algorithm NOUN bracis-19048 212 27 , , PUNCT bracis-19048 212 28 represent represent VERB bracis-19048 212 29 over over ADP bracis-19048 212 30 half half NOUN bracis-19048 212 31 of of ADP bracis-19048 212 32 the the DET bracis-19048 212 33 top top ADJ bracis-19048 212 34 features feature NOUN bracis-19048 212 35 . . PUNCT bracis-19048 213 1 interestingly interestingly ADV bracis-19048 213 2 , , PUNCT bracis-19048 213 3 six six NUM bracis-19048 213 4 of of ADP bracis-19048 213 5 the the DET bracis-19048 213 6 18 18 NUM bracis-19048 213 7 clustering clustering NOUN bracis-19048 213 8 features feature NOUN bracis-19048 213 9 are be AUX bracis-19048 213 10 ranked rank VERB bracis-19048 213 11 in in ADP bracis-19048 213 12 the the DET bracis-19048 213 13 20 20 NUM bracis-19048 213 14 most most ADV bracis-19048 213 15 relevant relevant ADJ bracis-19048 213 16 for for ADP bracis-19048 213 17 the the DET bracis-19048 213 18 rfr rfr PROPN bracis-19048 213 19 model model NOUN bracis-19048 213 20 , , PUNCT bracis-19048 213 21 in in ADP bracis-19048 213 22 this this DET bracis-19048 213 23 order order NOUN bracis-19048 213 24 : : PUNCT bracis-19048 213 25 sil sil PROPN bracis-19048 213 26 , , PUNCT bracis-19048 213 27 ch ch PROPN bracis-19048 213 28 , , PUNCT bracis-19048 213 29 ray ray PROPN bracis-19048 213 30 , , PUNCT bracis-19048 213 31 vdb vdb NOUN bracis-19048 213 32 , , PUNCT bracis-19048 213 33 c_index c_index NOUN bracis-19048 213 34 and and CCONJ bracis-19048 213 35 nre nre PROPN bracis-19048 213 36 . . PUNCT bracis-19048 214 1 this this PRON bracis-19048 214 2 suggests suggest VERB bracis-19048 214 3 that that SCONJ bracis-19048 214 4 these these DET bracis-19048 214 5 features feature NOUN bracis-19048 214 6 are be AUX bracis-19048 214 7 relevant relevant ADJ bracis-19048 214 8 to to ADP bracis-19048 214 9 the the DET bracis-19048 214 10 mtl mtl PROPN bracis-19048 214 11 task task NOUN bracis-19048 214 12 , , PUNCT bracis-19048 214 13 at at ADP bracis-19048 214 14 least least ADJ bracis-19048 214 15 for for ADP bracis-19048 214 16 the the DET bracis-19048 214 17 case case NOUN bracis-19048 214 18 of of ADP bracis-19048 214 19 the the DET bracis-19048 214 20 rfr rfr PROPN bracis-19048 214 21 meta meta PROPN bracis-19048 214 22 - - PUNCT bracis-19048 214 23 regressor regressor NOUN bracis-19048 214 24 , , PUNCT bracis-19048 214 25 which which PRON bracis-19048 214 26 would would AUX bracis-19048 214 27 corroborate corroborate VERB bracis-19048 214 28 the the DET bracis-19048 214 29 result result NOUN bracis-19048 214 30 of of ADP bracis-19048 214 31 the the DET bracis-19048 214 32 hypothesis hypothesis NOUN bracis-19048 214 33 test test NOUN bracis-19048 214 34 performed perform VERB bracis-19048 214 35 on on ADP bracis-19048 214 36 the the DET bracis-19048 214 37 data datum NOUN bracis-19048 214 38 of of ADP bracis-19048 214 39 fig fig NOUN bracis-19048 214 40 . . PUNCT bracis-19048 214 41     SPACE bracis-19048 215 1 2 2 X bracis-19048 215 2 . . PUNCT bracis-19048 215 3 finally finally ADV bracis-19048 215 4 , , PUNCT bracis-19048 215 5 the the DET bracis-19048 215 6 practical practical ADJ bracis-19048 215 7 applications application NOUN bracis-19048 215 8 of of ADP bracis-19048 215 9 using use VERB bracis-19048 215 10 clustering cluster VERB bracis-19048 215 11 meta meta NOUN bracis-19048 215 12 - - PUNCT bracis-19048 215 13 features feature NOUN bracis-19048 215 14 for for ADP bracis-19048 215 15 a a DET bracis-19048 215 16 mtl mtl PROPN bracis-19048 215 17 recommender recommender NOUN bracis-19048 215 18 system system NOUN bracis-19048 215 19 must must AUX bracis-19048 215 20 consider consider VERB bracis-19048 215 21 the the DET bracis-19048 215 22 computational computational ADJ bracis-19048 215 23 costs cost NOUN bracis-19048 215 24 involved involve VERB bracis-19048 215 25 . . PUNCT bracis-19048 216 1 the the DET bracis-19048 216 2 trade trade NOUN bracis-19048 216 3 - - PUNCT bracis-19048 216 4 off off NOUN bracis-19048 216 5 between between ADP bracis-19048 216 6 the the DET bracis-19048 216 7 runtime runtime NOUN bracis-19048 216 8 of of ADP bracis-19048 216 9 the the DET bracis-19048 216 10 characterization characterization NOUN bracis-19048 216 11 process process NOUN bracis-19048 216 12 and and CCONJ bracis-19048 216 13 the the DET bracis-19048 216 14 evaluation evaluation NOUN bracis-19048 216 15 of of ADP bracis-19048 216 16 all all DET bracis-19048 216 17 alternative alternative ADJ bracis-19048 216 18 data datum NOUN bracis-19048 216 19 modeling modeling NOUN bracis-19048 216 20 algorithms algorithm NOUN bracis-19048 216 21 considered consider VERB bracis-19048 216 22 to to PART bracis-19048 216 23 solve solve VERB bracis-19048 216 24 the the DET bracis-19048 216 25 task task NOUN bracis-19048 216 26 under under ADP bracis-19048 216 27 study study NOUN bracis-19048 216 28 should should AUX bracis-19048 216 29 favor favor VERB bracis-19048 216 30 the the DET bracis-19048 216 31 former former ADJ bracis-19048 216 32 in in ADP bracis-19048 216 33 order order NOUN bracis-19048 216 34 to to PART bracis-19048 216 35 make make VERB bracis-19048 216 36 the the DET bracis-19048 216 37 recommendation recommendation NOUN bracis-19048 216 38 system system NOUN bracis-19048 216 39 useful useful ADJ bracis-19048 216 40 . . PUNCT bracis-19048 217 1 figure figure NOUN bracis-19048 217 2     SPACE bracis-19048 217 3 5 5 NUM bracis-19048 217 4 presents present VERB bracis-19048 217 5 the the DET bracis-19048 217 6 results result NOUN bracis-19048 217 7 of of ADP bracis-19048 217 8 a a DET bracis-19048 217 9 runtime runtime NOUN bracis-19048 217 10 analysis analysis NOUN bracis-19048 217 11 to to PART bracis-19048 217 12 evaluate evaluate VERB bracis-19048 217 13 this this DET bracis-19048 217 14 trade trade NOUN bracis-19048 217 15 - - PUNCT bracis-19048 217 16 off off NOUN bracis-19048 217 17 for for ADP bracis-19048 217 18 the the DET bracis-19048 217 19 proposed propose VERB bracis-19048 217 20 approach approach NOUN bracis-19048 217 21 . . PUNCT bracis-19048 218 1 this this DET bracis-19048 218 2 analysis analysis NOUN bracis-19048 218 3 exhaustively exhaustively ADV bracis-19048 218 4 compared compare VERB bracis-19048 218 5 the the DET bracis-19048 218 6 single single ADJ bracis-19048 218 7 - - PUNCT bracis-19048 218 8 thread thread NOUN bracis-19048 218 9 runtime runtime NOUN bracis-19048 218 10 cost cost NOUN bracis-19048 218 11 for for ADP bracis-19048 218 12 extraction extraction NOUN bracis-19048 218 13 of of ADP bracis-19048 218 14 the the DET bracis-19048 218 15 standard standard NOUN bracis-19048 218 16 and and CCONJ bracis-19048 218 17 clustering cluster VERB bracis-19048 218 18 meta meta NOUN bracis-19048 218 19 - - PUNCT bracis-19048 218 20 features feature NOUN bracis-19048 218 21 to to ADP bracis-19048 218 22 the the DET bracis-19048 218 23 cost cost NOUN bracis-19048 218 24 for for ADP bracis-19048 218 25 running run VERB bracis-19048 218 26 all all DET bracis-19048 218 27 classification classification NOUN bracis-19048 218 28 algorithms algorithm NOUN bracis-19048 218 29 . . PUNCT bracis-19048 219 1 in in ADP bracis-19048 219 2 the the DET bracis-19048 219 3 figure figure NOUN bracis-19048 219 4 , , PUNCT bracis-19048 219 5 each each DET bracis-19048 219 6 point point NOUN bracis-19048 219 7 represents represent VERB bracis-19048 219 8 the the DET bracis-19048 219 9 cost cost NOUN bracis-19048 219 10 for for ADP bracis-19048 219 11 processing process VERB bracis-19048 219 12 a a DET bracis-19048 219 13 dataset dataset NOUN bracis-19048 219 14 in in ADP bracis-19048 219 15 a a DET bracis-19048 219 16 log\(\times log\(\time NOUN bracis-19048 219 17 \)log \)log NOUN bracis-19048 219 18 scale scale NOUN bracis-19048 219 19 . . PUNCT bracis-19048 220 1 the the DET bracis-19048 220 2 x x ADJ bracis-19048 220 3 - - ADJ bracis-19048 220 4 axis axis NOUN bracis-19048 220 5 shows show VERB bracis-19048 220 6 the the DET bracis-19048 220 7 cost cost NOUN bracis-19048 220 8 for for ADP bracis-19048 220 9 running run VERB bracis-19048 220 10 all all DET bracis-19048 220 11 classifiers classifier NOUN bracis-19048 220 12 while while SCONJ bracis-19048 220 13 the the DET bracis-19048 220 14 y y NOUN bracis-19048 220 15 - - PUNCT bracis-19048 220 16 axis axis NOUN bracis-19048 220 17 shows show VERB bracis-19048 220 18 the the DET bracis-19048 220 19 cost cost NOUN bracis-19048 220 20 for for ADP bracis-19048 220 21 extracting extract VERB bracis-19048 220 22 the the DET bracis-19048 220 23 meta meta NOUN bracis-19048 220 24 - - PUNCT bracis-19048 220 25 features feature NOUN bracis-19048 220 26 . . PUNCT bracis-19048 221 1 this this PRON bracis-19048 221 2 enables enable VERB bracis-19048 221 3 a a DET bracis-19048 221 4 straightforward straightforward ADJ bracis-19048 221 5 visual visual ADJ bracis-19048 221 6 analysis analysis NOUN bracis-19048 221 7 : : PUNCT bracis-19048 221 8 if if SCONJ bracis-19048 221 9 a a DET bracis-19048 221 10 point point NOUN bracis-19048 221 11 is be AUX bracis-19048 221 12 above above ADP bracis-19048 221 13 the the DET bracis-19048 221 14 line line NOUN bracis-19048 221 15 , , PUNCT bracis-19048 221 16 the the DET bracis-19048 221 17 y y PROPN bracis-19048 221 18 value value NOUN bracis-19048 221 19 is be AUX bracis-19048 221 20 less less ADJ bracis-19048 221 21 than than ADP bracis-19048 221 22 the the DET bracis-19048 221 23 x x NOUN bracis-19048 221 24 value value NOUN bracis-19048 221 25 , , PUNCT bracis-19048 221 26 thus thus ADV bracis-19048 221 27 running run VERB bracis-19048 221 28 the the DET bracis-19048 221 29 classifiers classifier NOUN bracis-19048 221 30 is be AUX bracis-19048 221 31 more more ADV bracis-19048 221 32 expensive expensive ADJ bracis-19048 221 33 than than ADP bracis-19048 221 34 extracting extract VERB bracis-19048 221 35 the the DET bracis-19048 221 36 features feature NOUN bracis-19048 221 37 . . PUNCT bracis-19048 222 1 fig fig NOUN bracis-19048 222 2 . . PUNCT bracis-19048 223 1 5 5 X bracis-19048 223 2 . . X bracis-19048 223 3 execution execution NOUN bracis-19048 223 4 times time NOUN bracis-19048 223 5 for for ADP bracis-19048 223 6 computing compute VERB bracis-19048 223 7 meta meta NOUN bracis-19048 223 8 - - PUNCT bracis-19048 223 9 features feature NOUN bracis-19048 223 10 compared compare VERB bracis-19048 223 11 to to ADP bracis-19048 223 12 applying apply VERB bracis-19048 223 13 all all DET bracis-19048 223 14 classifiers classifier NOUN bracis-19048 223 15 . . PUNCT bracis-19048 224 1 full full ADJ bracis-19048 224 2 size size NOUN bracis-19048 224 3 image image NOUN bracis-19048 224 4 the the DET bracis-19048 224 5 standard standard ADJ bracis-19048 224 6 meta meta NOUN bracis-19048 224 7 - - PUNCT bracis-19048 224 8 features feature NOUN bracis-19048 224 9 are be AUX bracis-19048 224 10 cheaper cheap ADJ bracis-19048 224 11 than than ADP bracis-19048 224 12 running run VERB bracis-19048 224 13 all all DET bracis-19048 224 14 classification classification NOUN bracis-19048 224 15 algorithms algorithm NOUN bracis-19048 224 16 in in ADP bracis-19048 224 17 around around ADP bracis-19048 224 18 89 89 NUM bracis-19048 224 19 % % NOUN bracis-19048 224 20 of of ADP bracis-19048 224 21 the the DET bracis-19048 224 22 cases case NOUN bracis-19048 224 23 . . PUNCT bracis-19048 225 1 the the DET bracis-19048 225 2 clustering clustering ADJ bracis-19048 225 3 features feature NOUN bracis-19048 225 4 are be AUX bracis-19048 225 5 considerably considerably ADV bracis-19048 225 6 cheaper cheap ADJ bracis-19048 225 7 than than ADP bracis-19048 225 8 that that PRON bracis-19048 225 9 running run VERB bracis-19048 225 10 all all DET bracis-19048 225 11 classification classification NOUN bracis-19048 225 12 algorithms algorithm NOUN bracis-19048 225 13 . . PUNCT bracis-19048 226 1 this this DET bracis-19048 226 2 low low ADJ bracis-19048 226 3 cost cost NOUN bracis-19048 226 4 in in ADP bracis-19048 226 5 computation computation NOUN bracis-19048 226 6 , , PUNCT bracis-19048 226 7 associated associate VERB bracis-19048 226 8 with with ADP bracis-19048 226 9 the the DET bracis-19048 226 10 informative informative ADJ bracis-19048 226 11 nature nature NOUN bracis-19048 226 12 of of ADP bracis-19048 226 13 some some PRON bracis-19048 226 14 of of ADP bracis-19048 226 15 the the DET bracis-19048 226 16 clustering clustering ADJ bracis-19048 226 17 meta meta NOUN bracis-19048 226 18 - - PUNCT bracis-19048 226 19 features feature NOUN bracis-19048 226 20 ( ( PUNCT bracis-19048 226 21 seen see VERB bracis-19048 226 22 in in ADP bracis-19048 226 23 fig fig NOUN bracis-19048 226 24 . . PUNCT bracis-19048 226 25     SPACE bracis-19048 226 26 4 4 NUM bracis-19048 226 27 ) ) PUNCT bracis-19048 226 28 , , PUNCT bracis-19048 226 29 suggests suggest VERB bracis-19048 226 30 their their PRON bracis-19048 226 31 usefulness usefulness NOUN bracis-19048 226 32 in in ADP bracis-19048 226 33 , , PUNCT bracis-19048 226 34 and and CCONJ bracis-19048 226 35 their their PRON bracis-19048 226 36 potential potential NOUN bracis-19048 226 37 for for ADP bracis-19048 226 38 , , PUNCT bracis-19048 226 39 improving improve VERB bracis-19048 226 40 mtl mtl PROPN bracis-19048 226 41 recommender recommender NOUN bracis-19048 226 42 systems system NOUN bracis-19048 226 43 . . PUNCT bracis-19048 227 1 5 5 NUM bracis-19048 227 2 conclusions conclusion NOUN bracis-19048 227 3 and and CCONJ bracis-19048 227 4     SPACE bracis-19048 227 5 future future NOUN bracis-19048 227 6 works work NOUN bracis-19048 227 7 this this DET bracis-19048 227 8 work work NOUN bracis-19048 227 9 investigated investigate VERB bracis-19048 227 10 both both CCONJ bracis-19048 227 11 the the DET bracis-19048 227 12 gains gain NOUN bracis-19048 227 13 in in ADP bracis-19048 227 14 performance performance NOUN bracis-19048 227 15 associated associate VERB bracis-19048 227 16 with with ADP bracis-19048 227 17 mtl mtl PROPN bracis-19048 227 18 in in ADP bracis-19048 227 19 general general ADJ bracis-19048 227 20 , , PUNCT bracis-19048 227 21 and and CCONJ bracis-19048 227 22 the the DET bracis-19048 227 23 use use NOUN bracis-19048 227 24 of of ADP bracis-19048 227 25 clustering cluster VERB bracis-19048 227 26 meta meta NOUN bracis-19048 227 27 - - PUNCT bracis-19048 227 28 features feature NOUN bracis-19048 227 29 in in ADP bracis-19048 227 30 a a DET bracis-19048 227 31 recommender recommender NOUN bracis-19048 227 32 system system NOUN bracis-19048 227 33 for for ADP bracis-19048 227 34 classification classification NOUN bracis-19048 227 35 algorithms algorithm NOUN bracis-19048 227 36 . . PUNCT bracis-19048 228 1 experimental experimental ADJ bracis-19048 228 2 results result NOUN bracis-19048 228 3 showed show VERB bracis-19048 228 4 that that SCONJ bracis-19048 228 5 the the DET bracis-19048 228 6 selected select VERB bracis-19048 228 7 classifiers classifier NOUN bracis-19048 228 8 were be AUX bracis-19048 228 9 adequate adequate ADJ bracis-19048 228 10 for for ADP bracis-19048 228 11 a a DET bracis-19048 228 12 recommender recommender NOUN bracis-19048 228 13 system system NOUN bracis-19048 228 14 and and CCONJ bracis-19048 228 15 that that SCONJ bracis-19048 228 16 using use VERB bracis-19048 228 17 the the DET bracis-19048 228 18 recommended recommend VERB bracis-19048 228 19 model model NOUN bracis-19048 228 20 resulted result VERB bracis-19048 228 21 in in ADP bracis-19048 228 22 improvements improvement NOUN bracis-19048 228 23 over over ADP bracis-19048 228 24 a a DET bracis-19048 228 25 random random ADJ bracis-19048 228 26 or or CCONJ bracis-19048 228 27 fixed fix VERB bracis-19048 228 28 ( ( PUNCT bracis-19048 228 29 best well ADV bracis-19048 228 30 expected expect VERB bracis-19048 228 31 ) ) PUNCT bracis-19048 228 32 algorithm algorithm NOUN bracis-19048 228 33 choice choice NOUN bracis-19048 228 34 . . PUNCT bracis-19048 229 1 the the DET bracis-19048 229 2 results result NOUN bracis-19048 229 3 presented present VERB bracis-19048 229 4 in in ADP bracis-19048 229 5 this this DET bracis-19048 229 6 paper paper NOUN bracis-19048 229 7 suggest suggest VERB bracis-19048 229 8 two two NUM bracis-19048 229 9 tentative tentative ADJ bracis-19048 229 10 conclusions conclusion NOUN bracis-19048 229 11 : : PUNCT bracis-19048 229 12 ( ( PUNCT bracis-19048 229 13 i i NOUN bracis-19048 229 14 ) ) PUNCT bracis-19048 229 15 that that PRON bracis-19048 229 16 meta meta NOUN bracis-19048 229 17 - - PUNCT bracis-19048 229 18 learning learning NOUN bracis-19048 229 19 , , PUNCT bracis-19048 229 20 as as ADP bracis-19048 229 21 an an DET bracis-19048 229 22 approach approach NOUN bracis-19048 229 23 to to PART bracis-19048 229 24 recommend recommend VERB bracis-19048 229 25 classifiers classifier NOUN bracis-19048 229 26 for for ADP bracis-19048 229 27 unseen unseen ADJ bracis-19048 229 28 problems problem NOUN bracis-19048 229 29 , , PUNCT bracis-19048 229 30 has have VERB bracis-19048 229 31 the the DET bracis-19048 229 32 potential potential NOUN bracis-19048 229 33 to to PART bracis-19048 229 34 provide provide VERB bracis-19048 229 35 good good ADJ bracis-19048 229 36 choices choice NOUN bracis-19048 229 37 with with ADP bracis-19048 229 38 a a DET bracis-19048 229 39 reduced reduced ADJ bracis-19048 229 40 computational computational ADJ bracis-19048 229 41 budget budget NOUN bracis-19048 229 42 ; ; PUNCT bracis-19048 229 43 and and CCONJ bracis-19048 229 44 ( ( PUNCT bracis-19048 229 45 ii ii NOUN bracis-19048 229 46 ) ) PUNCT bracis-19048 229 47 that that SCONJ bracis-19048 229 48 clustering clustering NOUN bracis-19048 229 49 - - PUNCT bracis-19048 229 50 based base VERB bracis-19048 229 51 meta meta NOUN bracis-19048 229 52 - - PUNCT bracis-19048 229 53 features feature NOUN bracis-19048 229 54 are be AUX bracis-19048 229 55 suitable suitable ADJ bracis-19048 229 56 to to PART bracis-19048 229 57 enhance enhance VERB bracis-19048 229 58 this this DET bracis-19048 229 59 mtl mtl PROPN bracis-19048 229 60 task task NOUN bracis-19048 229 61 , , PUNCT bracis-19048 229 62 which which PRON bracis-19048 229 63 may may AUX bracis-19048 229 64 indicate indicate VERB bracis-19048 229 65 that that SCONJ bracis-19048 229 66 they they PRON bracis-19048 229 67 are be AUX bracis-19048 229 68 able able ADJ bracis-19048 229 69 to to PART bracis-19048 229 70 capture capture VERB bracis-19048 229 71 relevant relevant ADJ bracis-19048 229 72 properties property NOUN bracis-19048 229 73 from from ADP bracis-19048 229 74 datasets dataset NOUN bracis-19048 229 75 to to PART bracis-19048 229 76 describe describe VERB bracis-19048 229 77 the the DET bracis-19048 229 78 performance performance NOUN bracis-19048 229 79 of of ADP bracis-19048 229 80 classifiers classifier NOUN bracis-19048 229 81 . . PUNCT bracis-19048 230 1 future future ADJ bracis-19048 230 2 works work NOUN bracis-19048 230 3 include include VERB bracis-19048 230 4 investigating investigate VERB bracis-19048 230 5 the the DET bracis-19048 230 6 impact impact NOUN bracis-19048 230 7 of of ADP bracis-19048 230 8 the the DET bracis-19048 230 9 underlying underlie VERB bracis-19048 230 10 grouping group VERB bracis-19048 230 11 structure structure NOUN bracis-19048 230 12 of of ADP bracis-19048 230 13 the the DET bracis-19048 230 14 datasets dataset NOUN bracis-19048 230 15 in in ADP bracis-19048 230 16 the the DET bracis-19048 230 17 adequacy adequacy NOUN bracis-19048 230 18 of of ADP bracis-19048 230 19 the the DET bracis-19048 230 20 clustering clustering ADJ bracis-19048 230 21 meta meta NOUN bracis-19048 230 22 - - PUNCT bracis-19048 230 23 features feature NOUN bracis-19048 230 24 , , PUNCT bracis-19048 230 25 as as ADV bracis-19048 230 26 well well ADV bracis-19048 230 27 as as ADP bracis-19048 230 28 investigating investigate VERB bracis-19048 230 29 the the DET bracis-19048 230 30 correlations correlation NOUN bracis-19048 230 31 of of ADP bracis-19048 230 32 the the DET bracis-19048 230 33 meta meta NOUN bracis-19048 230 34 - - PUNCT bracis-19048 230 35 features feature NOUN bracis-19048 230 36 to to PART bracis-19048 230 37 enable enable VERB bracis-19048 230 38 the the DET bracis-19048 230 39 selection selection NOUN bracis-19048 230 40 of of ADP bracis-19048 230 41 a a DET bracis-19048 230 42 more more ADV bracis-19048 230 43 efficient efficient ADJ bracis-19048 230 44 , , PUNCT bracis-19048 230 45 parsimonious parsimonious ADJ bracis-19048 230 46 subset subset NOUN bracis-19048 230 47 of of ADP bracis-19048 230 48 meta meta ADJ bracis-19048 230 49 - - PUNCT bracis-19048 230 50 feature feature NOUN bracis-19048 230 51 . . PUNCT bracis-19048 231 1 the the DET bracis-19048 231 2 incorporation incorporation NOUN bracis-19048 231 3 of of ADP bracis-19048 231 4 a a DET bracis-19048 231 5 more more ADV bracis-19048 231 6 diverse diverse ADJ bracis-19048 231 7 set set NOUN bracis-19048 231 8 of of ADP bracis-19048 231 9 classification classification NOUN bracis-19048 231 10 algorithms algorithm NOUN bracis-19048 231 11 for for ADP bracis-19048 231 12 the the DET bracis-19048 231 13 recommender recommender NOUN bracis-19048 231 14 system system NOUN bracis-19048 231 15 and and CCONJ bracis-19048 231 16 the the DET bracis-19048 231 17 incorporation incorporation NOUN bracis-19048 231 18 of of ADP bracis-19048 231 19 hyper hyper ADJ bracis-19048 231 20 - - ADJ bracis-19048 231 21 parameter parameter NOUN bracis-19048 231 22 tuning tuning NOUN bracis-19048 231 23 for for ADP bracis-19048 231 24 the the DET bracis-19048 231 25 classifiers classifier NOUN bracis-19048 231 26 and and CCONJ bracis-19048 231 27 regressors regressor NOUN bracis-19048 231 28 also also ADV bracis-19048 231 29 represent represent VERB bracis-19048 231 30 interesting interesting ADJ bracis-19048 231 31 new new ADJ bracis-19048 231 32 areas area NOUN bracis-19048 231 33 for for ADP bracis-19048 231 34 research research NOUN bracis-19048 231 35 . . PUNCT bracis-19048 232 1 notes note NOUN bracis-19048 232 2 1.https://github.com/rivolli/mfe 1.https://github.com/rivolli/mfe NUM bracis-19048 232 3 . . PUNCT bracis-19048 233 1 references reference NOUN bracis-19048 233 2 baker baker PROPN bracis-19048 233 3 , , PUNCT bracis-19048 233 4 f.b f.b PROPN bracis-19048 233 5 . . PROPN bracis-19048 233 6 , , PUNCT bracis-19048 233 7 hubert hubert PROPN bracis-19048 233 8 , , PUNCT bracis-19048 233 9 l.j l.j PROPN bracis-19048 233 10 . . PROPN bracis-19048 233 11 : : PUNCT bracis-19048 233 12 measuring measure VERB bracis-19048 233 13 the the DET bracis-19048 233 14 power power NOUN bracis-19048 233 15 of of ADP bracis-19048 233 16 hierarchical hierarchical ADJ bracis-19048 233 17 cluster cluster NOUN bracis-19048 233 18 analysis analysis NOUN bracis-19048 233 19 . . PUNCT bracis-19048 234 1 j. j. PROPN bracis-19048 234 2 am am PROPN bracis-19048 234 3 . . PUNCT bracis-19048 235 1 stat stat PROPN bracis-19048 235 2 . . PUNCT bracis-19048 236 1 assoc assoc PROPN bracis-19048 236 2 . . PUNCT bracis-19048 237 1 70(349 70(349 NOUN bracis-19048 237 2 ) ) PUNCT bracis-19048 237 3 , , PUNCT bracis-19048 237 4 31–38 31–38 NUM bracis-19048 237 5 ( ( PUNCT bracis-19048 237 6 1975 1975 NUM bracis-19048 237 7 ) ) PUNCT bracis-19048 237 8 article article NOUN bracis-19048 237 9     SPACE bracis-19048 237 10 google google PROPN bracis-19048 237 11 scholar scholar NOUN bracis-19048 237 12     SPACE bracis-19048 237 13 bezdek bezdek PROPN bracis-19048 237 14 , , PUNCT bracis-19048 237 15 j.c j.c PROPN bracis-19048 237 16 . . PROPN bracis-19048 237 17 , , PUNCT bracis-19048 237 18 pal pal NOUN bracis-19048 237 19 , , PUNCT bracis-19048 237 20 n.r n.r PROPN bracis-19048 237 21 . . PROPN bracis-19048 237 22 : : PUNCT bracis-19048 238 1 some some DET bracis-19048 238 2 new new ADJ bracis-19048 238 3 indexes index NOUN bracis-19048 238 4 of of ADP bracis-19048 238 5 cluster cluster NOUN bracis-19048 238 6 validity validity NOUN bracis-19048 238 7 . . PUNCT bracis-19048 239 1 ieee ieee PROPN bracis-19048 239 2 trans trans PROPN bracis-19048 239 3 . . PUNCT bracis-19048 240 1 syst syst PROPN bracis-19048 240 2 . . PUNCT bracis-19048 241 1 man man PROPN bracis-19048 241 2 cybern cybern PROPN bracis-19048 241 3 . . PUNCT bracis-19048 242 1 part part NOUN bracis-19048 242 2 b b PROPN bracis-19048 242 3 ( ( PUNCT bracis-19048 242 4 cybern cybern NOUN bracis-19048 242 5 . . PUNCT bracis-19048 242 6 ) ) PUNCT bracis-19048 243 1 28(3 28(3 NUM bracis-19048 243 2 ) ) PUNCT bracis-19048 243 3 , , PUNCT bracis-19048 243 4 301–315 301–315 NUM bracis-19048 243 5 ( ( PUNCT bracis-19048 243 6 1998 1998 NUM bracis-19048 243 7 ) ) PUNCT bracis-19048 243 8 google google PROPN bracis-19048 243 9 scholar scholar NOUN bracis-19048 243 10     SPACE bracis-19048 243 11 bilalli bilalli PROPN bracis-19048 243 12 , , PUNCT bracis-19048 243 13 b. b. PROPN bracis-19048 243 14 , , PUNCT bracis-19048 243 15 abelló abelló PROPN bracis-19048 243 16 , , PUNCT bracis-19048 243 17 a. a. NOUN bracis-19048 243 18 , , PUNCT bracis-19048 243 19 aluja aluja NOUN bracis-19048 243 20 - - PUNCT bracis-19048 243 21 banet banet NOUN bracis-19048 243 22 , , PUNCT bracis-19048 243 23 t. t. PROPN bracis-19048 243 24 : : PUNCT bracis-19048 243 25 on on ADP bracis-19048 243 26 the the DET bracis-19048 243 27 predictive predictive ADJ bracis-19048 243 28 power power NOUN bracis-19048 243 29 of of ADP bracis-19048 243 30 meta meta NOUN bracis-19048 243 31 - - PUNCT bracis-19048 243 32 features feature NOUN bracis-19048 243 33 in in ADP bracis-19048 243 34 openml openml NOUN bracis-19048 243 35 . . PUNCT bracis-19048 244 1 int int NOUN bracis-19048 244 2 . . PUNCT bracis-19048 245 1 j. j. PROPN bracis-19048 245 2 appl appl PROPN bracis-19048 245 3 . . PROPN bracis-19048 245 4 math math PROPN bracis-19048 245 5 . . PUNCT bracis-19048 246 1 comput comput NOUN bracis-19048 246 2 . . PUNCT bracis-19048 247 1 sci sci PROPN bracis-19048 247 2 . . PUNCT bracis-19048 248 1 27(4 27(4 PROPN bracis-19048 248 2 ) ) PUNCT bracis-19048 248 3 , , PUNCT bracis-19048 248 4 697–712 697–712 NUM bracis-19048 248 5 ( ( PUNCT bracis-19048 248 6 2017 2017 NUM bracis-19048 248 7 ) ) PUNCT bracis-19048 248 8 article article NOUN bracis-19048 248 9     SPACE bracis-19048 248 10 mathscinet mathscinet NOUN bracis-19048 248 11     SPACE bracis-19048 248 12 google google PROPN bracis-19048 248 13 scholar scholar NOUN bracis-19048 248 14     SPACE bracis-19048 248 15 brazdil brazdil NOUN bracis-19048 248 16 , , PUNCT bracis-19048 248 17 p. p. NOUN bracis-19048 248 18 , , PUNCT bracis-19048 248 19 giraud giraud NOUN bracis-19048 248 20 - - PUNCT bracis-19048 248 21 carrier carrier NOUN bracis-19048 248 22 , , PUNCT bracis-19048 248 23 c. c. PROPN bracis-19048 248 24 , , PUNCT bracis-19048 248 25 soares soares PROPN bracis-19048 248 26 , , PUNCT bracis-19048 248 27 c. c. PROPN bracis-19048 248 28 , , PUNCT bracis-19048 248 29 vilalta vilalta PROPN bracis-19048 248 30 , , PUNCT bracis-19048 248 31 r. r. PROPN bracis-19048 248 32 : : PUNCT bracis-19048 248 33 metalearning metalearne VERB bracis-19048 248 34 applications application NOUN bracis-19048 248 35 to to ADP bracis-19048 248 36 data datum NOUN bracis-19048 248 37 mining mining NOUN bracis-19048 248 38 . . PUNCT bracis-19048 249 1 cognitive cognitive ADJ bracis-19048 249 2 technologies technology NOUN bracis-19048 249 3 , , PUNCT bracis-19048 249 4 1st 1st ADJ bracis-19048 249 5 edn edn PROPN bracis-19048 249 6 . . PUNCT bracis-19048 249 7 springer springer NOUN bracis-19048 249 8 , , PUNCT bracis-19048 249 9 heidelberg heidelberg PROPN bracis-19048 249 10 ( ( PUNCT bracis-19048 249 11 2009 2009 NUM bracis-19048 249 12 ) ) PUNCT bracis-19048 249 13 . . PUNCT bracis-19048 250 1 https://doi.org/10.1007/978-3-540-73263-1 https://doi.org/10.1007/978-3-540-73263-1 PROPN bracis-19048 250 2 book book NOUN bracis-19048 250 3     SPACE bracis-19048 250 4 math math NOUN bracis-19048 250 5     SPACE bracis-19048 250 6 google google PROPN bracis-19048 250 7 scholar scholar NOUN bracis-19048 250 8     SPACE bracis-19048 250 9 breiman breiman NOUN bracis-19048 250 10 , , PUNCT bracis-19048 250 11 l. l. PROPN bracis-19048 250 12 : : PUNCT bracis-19048 250 13 random random ADJ bracis-19048 250 14 forests forest NOUN bracis-19048 250 15 . . PUNCT bracis-19048 251 1 mach mach NOUN bracis-19048 251 2 . . PUNCT bracis-19048 252 1 learn learn PROPN bracis-19048 252 2 . . PUNCT bracis-19048 253 1 45(1 45(1 NOUN bracis-19048 253 2 ) ) PUNCT bracis-19048 253 3 , , PUNCT bracis-19048 253 4 5–32 5–32 NOUN bracis-19048 253 5 ( ( PUNCT bracis-19048 253 6 2001 2001 NUM bracis-19048 253 7 ) ) PUNCT bracis-19048 253 8 article article NOUN bracis-19048 253 9     SPACE bracis-19048 253 10 google google PROPN bracis-19048 253 11 scholar scholar NOUN bracis-19048 253 12     SPACE bracis-19048 253 13 breiman breiman NOUN bracis-19048 253 14 , , PUNCT bracis-19048 253 15 l. l. PROPN bracis-19048 253 16 , , PUNCT bracis-19048 253 17 friedman friedman PROPN bracis-19048 253 18 , , PUNCT bracis-19048 253 19 j.h j.h PROPN bracis-19048 253 20 . . PROPN bracis-19048 253 21 , , PUNCT bracis-19048 253 22 olshen olshen NOUN bracis-19048 253 23 , , PUNCT bracis-19048 253 24 r.a r.a PROPN bracis-19048 253 25 . . PROPN bracis-19048 253 26 , , PUNCT bracis-19048 253 27 stone stone NOUN bracis-19048 253 28 , , PUNCT bracis-19048 253 29 c.j c.j PROPN bracis-19048 253 30 . . NOUN bracis-19048 253 31 : : PUNCT bracis-19048 254 1 classification classification NOUN bracis-19048 254 2 and and CCONJ bracis-19048 254 3 regression regression NOUN bracis-19048 254 4 trees tree NOUN bracis-19048 254 5 . . PUNCT bracis-19048 255 1 wadsworth wadsworth NOUN bracis-19048 255 2 and and CCONJ bracis-19048 255 3 brooks brooks PROPN bracis-19048 255 4 ( ( PUNCT bracis-19048 255 5 1984 1984 NUM bracis-19048 255 6 ) ) PUNCT bracis-19048 255 7 google google NOUN bracis-19048 255 8 scholar scholar NOUN bracis-19048 255 9     SPACE bracis-19048 255 10 brock brock PROPN bracis-19048 255 11 , , PUNCT bracis-19048 255 12 g. g. PROPN bracis-19048 255 13 , , PUNCT bracis-19048 255 14 pihur pihur VERB bracis-19048 255 15 , , PUNCT bracis-19048 255 16 v. v. PROPN bracis-19048 255 17 , , PUNCT bracis-19048 255 18 datta datta PROPN bracis-19048 255 19 , , PUNCT bracis-19048 255 20 s. s. PROPN bracis-19048 255 21 , , PUNCT bracis-19048 255 22 datta datta PROPN bracis-19048 255 23 , , PUNCT bracis-19048 255 24 s. s. PROPN bracis-19048 255 25 : : PUNCT bracis-19048 256 1 clvalid clvalid PROPN bracis-19048 256 2 : : PUNCT bracis-19048 256 3 an an DET bracis-19048 256 4 r r NOUN bracis-19048 256 5 package package NOUN bracis-19048 256 6 for for ADP bracis-19048 256 7 cluster cluster NOUN bracis-19048 256 8 validation validation NOUN bracis-19048 256 9 . . PUNCT bracis-19048 257 1 j. j. PROPN bracis-19048 257 2 stat stat PROPN bracis-19048 257 3 . . PUNCT bracis-19048 258 1 softw softw PROPN bracis-19048 258 2 . . PUNCT bracis-19048 259 1 25(4 25(4 NOUN bracis-19048 259 2 ) ) PUNCT bracis-19048 259 3 , , PUNCT bracis-19048 259 4 1–22 1–22 PROPN bracis-19048 259 5 ( ( PUNCT bracis-19048 259 6 2008 2008 NUM bracis-19048 259 7 ) ) PUNCT bracis-19048 259 8 . . PUNCT bracis-19048 260 1 http://www.jstatsoft.org/v25/i04/ http://www.jstatsoft.org/v25/i04/ PROPN bracis-19048 261 1 caliński caliński NOUN bracis-19048 261 2 , , PUNCT bracis-19048 261 3 t. t. PROPN bracis-19048 261 4 , , PUNCT bracis-19048 261 5 harabasz harabasz PROPN bracis-19048 261 6 , , PUNCT bracis-19048 261 7 j. j. PROPN bracis-19048 261 8 : : PUNCT bracis-19048 261 9 a a DET bracis-19048 261 10 dendrite dendrite ADJ bracis-19048 261 11 method method NOUN bracis-19048 261 12 for for ADP bracis-19048 261 13 cluster cluster NOUN bracis-19048 261 14 analysis analysis NOUN bracis-19048 261 15 . . PUNCT bracis-19048 262 1 commun commun PROPN bracis-19048 262 2 . . PUNCT bracis-19048 263 1 stat stat PROPN bracis-19048 263 2 . . PUNCT bracis-19048 264 1 theory theory NOUN bracis-19048 264 2 methods method NOUN bracis-19048 264 3 3(1 3(1 NUM bracis-19048 264 4 ) ) PUNCT bracis-19048 264 5 , , PUNCT bracis-19048 264 6 1–27 1–27 PROPN bracis-19048 264 7 ( ( PUNCT bracis-19048 264 8 1974 1974 NUM bracis-19048 264 9 ) ) PUNCT bracis-19048 264 10 article article NOUN bracis-19048 264 11     SPACE bracis-19048 264 12 mathscinet mathscinet NOUN bracis-19048 264 13     SPACE bracis-19048 264 14 google google PROPN bracis-19048 264 15 scholar scholar NOUN bracis-19048 264 16     SPACE bracis-19048 264 17 castiello castiello PROPN bracis-19048 264 18 , , PUNCT bracis-19048 264 19 c. c. PROPN bracis-19048 264 20 , , PUNCT bracis-19048 264 21 castellano castellano NOUN bracis-19048 264 22 , , PUNCT bracis-19048 264 23 g. g. PROPN bracis-19048 264 24 , , PUNCT bracis-19048 264 25 fanelli fanelli PROPN bracis-19048 264 26 , , PUNCT bracis-19048 264 27 a.m. a.m. ADV bracis-19048 264 28 : : PUNCT bracis-19048 264 29 meta meta ADJ bracis-19048 264 30 - - PUNCT bracis-19048 264 31 data data NOUN bracis-19048 264 32 : : PUNCT bracis-19048 264 33 characterization characterization NOUN bracis-19048 264 34 of of ADP bracis-19048 264 35 input input NOUN bracis-19048 264 36 features feature NOUN bracis-19048 264 37 for for ADP bracis-19048 264 38 meta meta NOUN bracis-19048 264 39 - - PUNCT bracis-19048 264 40 learning learning NOUN bracis-19048 264 41 . . PUNCT bracis-19048 265 1 in in ADP bracis-19048 265 2 : : PUNCT bracis-19048 265 3 torra torra PROPN bracis-19048 265 4 , , PUNCT bracis-19048 265 5 v. v. ADV bracis-19048 265 6 , , PUNCT bracis-19048 265 7 narukawa narukawa PROPN bracis-19048 265 8 , , PUNCT bracis-19048 265 9 y. y. PROPN bracis-19048 265 10 , , PUNCT bracis-19048 265 11 miyamoto miyamoto NOUN bracis-19048 265 12 , , PUNCT bracis-19048 265 13 s. s. PROPN bracis-19048 265 14 ( ( PUNCT bracis-19048 265 15 eds eds PROPN bracis-19048 265 16 . . PUNCT bracis-19048 265 17 ) ) PUNCT bracis-19048 266 1 mdai mdai PROPN bracis-19048 266 2 2005 2005 NUM bracis-19048 266 3 . . PUNCT bracis-19048 267 1 lncs lncs PROPN bracis-19048 267 2 ( ( PUNCT bracis-19048 267 3 lnai lnai ADJ bracis-19048 267 4 ) ) PUNCT bracis-19048 267 5 , , PUNCT bracis-19048 267 6 vol vol NOUN bracis-19048 267 7 . . PROPN bracis-19048 267 8 3558 3558 NUM bracis-19048 267 9 , , PUNCT bracis-19048 267 10 pp pp ADV bracis-19048 267 11 . . PUNCT bracis-19048 268 1 457–468 457–468 NUM bracis-19048 268 2 . . PUNCT bracis-19048 268 3 springer springer PROPN bracis-19048 268 4 , , PUNCT bracis-19048 268 5 heidelberg heidelberg PROPN bracis-19048 268 6 ( ( PUNCT bracis-19048 268 7 2005 2005 NUM bracis-19048 268 8 ) ) PUNCT bracis-19048 268 9 . . PUNCT bracis-19048 269 1 https://doi.org/10.1007/11526018_45 https://doi.org/10.1007/11526018_45 PROPN bracis-19048 269 2 chapter chapter NOUN bracis-19048 269 3     SPACE bracis-19048 269 4 math math NOUN bracis-19048 269 5     SPACE bracis-19048 269 6 google google PROPN bracis-19048 269 7 scholar scholar NOUN bracis-19048 269 8     SPACE bracis-19048 269 9 cristianini cristianini PROPN bracis-19048 269 10 , , PUNCT bracis-19048 269 11 n. n. NOUN bracis-19048 269 12 , , PUNCT bracis-19048 269 13 shawe shawe NOUN bracis-19048 269 14 - - PUNCT bracis-19048 269 15 taylor taylor PROPN bracis-19048 269 16 , , PUNCT bracis-19048 269 17 j. j. PROPN bracis-19048 269 18 : : PUNCT bracis-19048 269 19 an an DET bracis-19048 269 20 introduction introduction NOUN bracis-19048 269 21 to to PART bracis-19048 269 22 support support VERB bracis-19048 269 23 vector vector NOUN bracis-19048 269 24 machines machine NOUN bracis-19048 269 25 and and CCONJ bracis-19048 269 26 other other ADJ bracis-19048 269 27 kernel kernel NOUN bracis-19048 269 28 - - PUNCT bracis-19048 269 29 based base VERB bracis-19048 269 30 learning learning NOUN bracis-19048 269 31 methods method NOUN bracis-19048 269 32 . . PUNCT bracis-19048 270 1 cambridge cambridge PROPN bracis-19048 270 2 university university PROPN bracis-19048 270 3 press press PROPN bracis-19048 270 4 , , PUNCT bracis-19048 270 5 cambridge cambridge PROPN bracis-19048 270 6 ( ( PUNCT bracis-19048 270 7 2000 2000 NUM bracis-19048 270 8 ) ) PUNCT bracis-19048 270 9 google google NOUN bracis-19048 270 10 scholar scholar NOUN bracis-19048 270 11     SPACE bracis-19048 270 12 davies davy NOUN bracis-19048 270 13 , , PUNCT bracis-19048 270 14 d.l d.l PROPN bracis-19048 270 15 . . PROPN bracis-19048 270 16 , , PUNCT bracis-19048 270 17 bouldin bouldin PROPN bracis-19048 270 18 , , PUNCT bracis-19048 270 19 d.w d.w PROPN bracis-19048 270 20 . . PROPN bracis-19048 270 21 : : PUNCT bracis-19048 270 22 a a DET bracis-19048 270 23 cluster cluster NOUN bracis-19048 270 24 separation separation NOUN bracis-19048 270 25 measure measure NOUN bracis-19048 270 26 . . PUNCT bracis-19048 271 1 ieee ieee PROPN bracis-19048 271 2 trans trans PROPN bracis-19048 271 3 . . PUNCT bracis-19048 271 4 pattern pattern PROPN bracis-19048 271 5 anal anal PROPN bracis-19048 271 6 . . PUNCT bracis-19048 272 1 mach mach PROPN bracis-19048 272 2 . . PUNCT bracis-19048 273 1 intell intell PROPN bracis-19048 273 2 . . PUNCT bracis-19048 274 1 pami-1(2 pami-1(2 X bracis-19048 274 2 ) ) PUNCT bracis-19048 274 3 , , PUNCT bracis-19048 275 1 224–227 224–227 NUM bracis-19048 275 2 ( ( PUNCT bracis-19048 275 3 1979 1979 NUM bracis-19048 275 4 ) ) PUNCT bracis-19048 275 5 google google NOUN bracis-19048 275 6 scholar scholar NOUN bracis-19048 275 7     SPACE bracis-19048 275 8 desgraupes desgraupe NOUN bracis-19048 275 9 , , PUNCT bracis-19048 275 10 b. b. NOUN bracis-19048 275 11 : : PUNCT bracis-19048 275 12 clustercrit clustercrit NOUN bracis-19048 275 13 vignette vignette NOUN bracis-19048 275 14 ( ( PUNCT bracis-19048 275 15 2018 2018 NUM bracis-19048 275 16 ) ) PUNCT bracis-19048 275 17 . . PUNCT bracis-19048 276 1 https://cran.r-project.org/package=clustercrit/vignettes/clustercrit.pdf https://cran.r-project.org/package=clustercrit/vignettes/clustercrit.pdf PUNCT bracis-19048 276 2 dua dua PROPN bracis-19048 276 3 , , PUNCT bracis-19048 276 4 d. d. PROPN bracis-19048 276 5 , , PUNCT bracis-19048 276 6 graff graff PROPN bracis-19048 276 7 , , PUNCT bracis-19048 276 8 c. c. PROPN bracis-19048 276 9 : : PUNCT bracis-19048 276 10 uci uci PROPN bracis-19048 276 11 machine machine NOUN bracis-19048 276 12 learning learn VERB bracis-19048 276 13 repository repository NOUN bracis-19048 276 14 ( ( PUNCT bracis-19048 276 15 2017 2017 NUM bracis-19048 276 16 ) ) PUNCT bracis-19048 276 17 . . PUNCT bracis-19048 277 1 http://archive.ics.uci.edu/ml http://archive.ics.uci.edu/ml NUM bracis-19048 277 2 dunn dunn PROPN bracis-19048 277 3 , , PUNCT bracis-19048 277 4 j.c j.c PROPN bracis-19048 277 5 . . PROPN bracis-19048 277 6 : : PUNCT bracis-19048 277 7 well well ADV bracis-19048 277 8 - - PUNCT bracis-19048 277 9 separated separate VERB bracis-19048 277 10 clusters cluster NOUN bracis-19048 277 11 and and CCONJ bracis-19048 277 12 optimal optimal ADJ bracis-19048 277 13 fuzzy fuzzy ADJ bracis-19048 277 14 partitions partition NOUN bracis-19048 277 15 . . PUNCT bracis-19048 278 1 j. j. PROPN bracis-19048 278 2 cybern cybern PROPN bracis-19048 278 3 . . PUNCT bracis-19048 279 1 4(1 4(1 NOUN bracis-19048 279 2 ) ) PUNCT bracis-19048 279 3 , , PUNCT bracis-19048 280 1 95–104 95–104 PROPN bracis-19048 280 2 ( ( PUNCT bracis-19048 280 3 1974 1974 NUM bracis-19048 280 4 ) ) PUNCT bracis-19048 280 5 article article NOUN bracis-19048 280 6     SPACE bracis-19048 280 7 mathscinet mathscinet NOUN bracis-19048 280 8     SPACE bracis-19048 280 9 google google PROPN bracis-19048 280 10 scholar scholar NOUN bracis-19048 280 11     SPACE bracis-19048 280 12 filchenkov filchenkov NOUN bracis-19048 280 13 , , PUNCT bracis-19048 280 14 a. a. NOUN bracis-19048 280 15 , , PUNCT bracis-19048 280 16 pendryak pendryak NOUN bracis-19048 280 17 , , PUNCT bracis-19048 280 18 a. a. NOUN bracis-19048 280 19 : : PUNCT bracis-19048 280 20 datasets dataset NOUN bracis-19048 280 21 meta meta ADJ bracis-19048 280 22 - - PUNCT bracis-19048 280 23 feature feature NOUN bracis-19048 280 24 description description NOUN bracis-19048 280 25 for for ADP bracis-19048 280 26 recommending recommend VERB bracis-19048 280 27 feature feature NOUN bracis-19048 280 28 selection selection NOUN bracis-19048 280 29 algorithm algorithm NOUN bracis-19048 280 30 . . PUNCT bracis-19048 281 1 in in ADP bracis-19048 281 2 : : PUNCT bracis-19048 281 3 artificial artificial ADJ bracis-19048 281 4 intelligence intelligence NOUN bracis-19048 281 5 and and CCONJ bracis-19048 281 6 natural natural ADJ bracis-19048 281 7 language language NOUN bracis-19048 281 8 and and CCONJ bracis-19048 281 9 information information NOUN bracis-19048 281 10 extraction extraction NOUN bracis-19048 281 11 , , PUNCT bracis-19048 281 12 social social ADJ bracis-19048 281 13 media medium NOUN bracis-19048 281 14 and and CCONJ bracis-19048 281 15 web web NOUN bracis-19048 281 16 search search NOUN bracis-19048 281 17 fruct fruct NOUN bracis-19048 281 18 conference conference NOUN bracis-19048 281 19 ( ( PUNCT bracis-19048 281 20 ainl ainl NOUN bracis-19048 281 21 - - PUNCT bracis-19048 281 22 ismw ismw NOUN bracis-19048 281 23 fruct fruct NOUN bracis-19048 281 24 ) ) PUNCT bracis-19048 281 25 , , PUNCT bracis-19048 281 26 vol vol NOUN bracis-19048 281 27 . . PROPN bracis-19048 281 28 7 7 NUM bracis-19048 281 29 , , PUNCT bracis-19048 281 30 pp pp ADJ bracis-19048 281 31 . . PUNCT bracis-19048 282 1 11–18 11–18 NUM bracis-19048 282 2 ( ( PUNCT bracis-19048 282 3 2015 2015 NUM bracis-19048 282 4 ) ) PUNCT bracis-19048 282 5 google google PROPN bracis-19048 282 6 scholar scholar NOUN bracis-19048 282 7     SPACE bracis-19048 282 8 garcia garcia PROPN bracis-19048 282 9 , , PUNCT bracis-19048 282 10 l.p.f l.p.f VERB bracis-19048 282 11 . . PROPN bracis-19048 282 12 , , PUNCT bracis-19048 282 13 lorena lorena PROPN bracis-19048 282 14 , , PUNCT bracis-19048 282 15 a.c a.c PROPN bracis-19048 282 16 . . PROPN bracis-19048 282 17 , , PUNCT bracis-19048 282 18 de de X bracis-19048 282 19 souto souto PROPN bracis-19048 282 20 , , PUNCT bracis-19048 282 21 m.c.p m.c.p NOUN bracis-19048 282 22 . . PROPN bracis-19048 282 23 , , PUNCT bracis-19048 282 24 ho ho PROPN bracis-19048 282 25 , , PUNCT bracis-19048 282 26 t.k t.k PROPN bracis-19048 282 27 . . PROPN bracis-19048 282 28 : : PUNCT bracis-19048 283 1 classifier classifier NOUN bracis-19048 283 2 recommendation recommendation NOUN bracis-19048 283 3 using use VERB bracis-19048 283 4 data datum NOUN bracis-19048 283 5 complexity complexity NOUN bracis-19048 283 6 measures measure NOUN bracis-19048 283 7 . . PUNCT bracis-19048 284 1 in in ADP bracis-19048 284 2 : : PUNCT bracis-19048 284 3 24th 24th ADJ bracis-19048 284 4 international international ADJ bracis-19048 284 5 conference conference NOUN bracis-19048 284 6 on on ADP bracis-19048 284 7 pattern pattern NOUN bracis-19048 284 8 recognition recognition NOUN bracis-19048 284 9 ( ( PUNCT bracis-19048 284 10 icpr icpr PROPN bracis-19048 284 11 ) ) PUNCT bracis-19048 284 12 , , PUNCT bracis-19048 284 13 pp pp ADP bracis-19048 284 14 . . PUNCT bracis-19048 285 1 874–879 874–879 NUM bracis-19048 285 2 ( ( PUNCT bracis-19048 285 3 2018 2018 NUM bracis-19048 285 4 ) ) PUNCT bracis-19048 285 5 google google PROPN bracis-19048 285 6 scholar scholar NOUN bracis-19048 285 7     SPACE bracis-19048 285 8 garcia garcia PROPN bracis-19048 285 9 , , PUNCT bracis-19048 285 10 l.p.f l.p.f VERB bracis-19048 285 11 . . PROPN bracis-19048 285 12 , , PUNCT bracis-19048 285 13 rivolli rivolli PROPN bracis-19048 285 14 , , PUNCT bracis-19048 285 15 a. a. NOUN bracis-19048 285 16 , , PUNCT bracis-19048 285 17 alcobaça alcobaça PROPN bracis-19048 285 18 , , PUNCT bracis-19048 285 19 e. e. PROPN bracis-19048 285 20 , , PUNCT bracis-19048 285 21 lorena lorena PROPN bracis-19048 285 22 , , PUNCT bracis-19048 285 23 a.c a.c PROPN bracis-19048 285 24 . . PROPN bracis-19048 285 25 , , PUNCT bracis-19048 285 26 de de X bracis-19048 285 27 carvalho carvalho NOUN bracis-19048 285 28 , , PUNCT bracis-19048 285 29 a.c.p.l.f a.c.p.l.f PROPN bracis-19048 285 30 . . PUNCT bracis-19048 285 31 : : PUNCT bracis-19048 286 1 boosting boost VERB bracis-19048 286 2 meta meta NOUN bracis-19048 286 3 - - PUNCT bracis-19048 286 4 learning learning NOUN bracis-19048 286 5 with with ADP bracis-19048 286 6 simulated simulated ADJ bracis-19048 286 7 data datum NOUN bracis-19048 286 8 complexity complexity NOUN bracis-19048 286 9 measures measure NOUN bracis-19048 286 10 . . PUNCT bracis-19048 287 1 intell intell PROPN bracis-19048 287 2 . . PUNCT bracis-19048 288 1 data data PROPN bracis-19048 288 2 anal anal PROPN bracis-19048 288 3 . . PUNCT bracis-19048 289 1 24(5 24(5 NUM bracis-19048 289 2 ) ) PUNCT bracis-19048 289 3 , , PUNCT bracis-19048 289 4 1011–1028 1011–1028 NUM bracis-19048 289 5 ( ( PUNCT bracis-19048 289 6 2020 2020 NUM bracis-19048 289 7 ) ) PUNCT bracis-19048 289 8 google google PROPN bracis-19048 289 9 scholar scholar NOUN bracis-19048 289 10     SPACE bracis-19048 289 11 halkidi halkidi PROPN bracis-19048 289 12 , , PUNCT bracis-19048 289 13 m. m. NOUN bracis-19048 289 14 , , PUNCT bracis-19048 289 15 batistakis batistakis PROPN bracis-19048 289 16 , , PUNCT bracis-19048 289 17 y. y. PROPN bracis-19048 289 18 , , PUNCT bracis-19048 289 19 vazirgiannis vazirgiannis PROPN bracis-19048 289 20 , , PUNCT bracis-19048 289 21 m. m. NOUN bracis-19048 289 22 : : PUNCT bracis-19048 289 23 on on ADP bracis-19048 289 24 clustering cluster VERB bracis-19048 289 25 validation validation NOUN bracis-19048 289 26 techniques technique NOUN bracis-19048 289 27 . . PUNCT bracis-19048 290 1 j. j. PROPN bracis-19048 290 2 intell intell PROPN bracis-19048 290 3 . . PUNCT bracis-19048 291 1 inf inf PROPN bracis-19048 291 2 . . PUNCT bracis-19048 291 3 syst syst PROPN bracis-19048 291 4 . . PUNCT bracis-19048 292 1 17(2–3 17(2–3 NUM bracis-19048 292 2 ) ) PUNCT bracis-19048 292 3 , , PUNCT bracis-19048 292 4 107–145 107–145 NUM bracis-19048 292 5 ( ( PUNCT bracis-19048 292 6 2001 2001 NUM bracis-19048 292 7 ) ) PUNCT bracis-19048 292 8 article article NOUN bracis-19048 292 9     SPACE bracis-19048 292 10 google google PROPN bracis-19048 292 11 scholar scholar NOUN bracis-19048 292 12     SPACE bracis-19048 292 13 handl handl PROPN bracis-19048 292 14 , , PUNCT bracis-19048 292 15 j. j. PROPN bracis-19048 292 16 , , PUNCT bracis-19048 292 17 knowles knowles PROPN bracis-19048 292 18 , , PUNCT bracis-19048 292 19 j. j. PROPN bracis-19048 292 20 : : PUNCT bracis-19048 292 21 exploiting exploit VERB bracis-19048 292 22 the the DET bracis-19048 292 23 trade trade NOUN bracis-19048 292 24 - - PUNCT bracis-19048 292 25 off off NOUN bracis-19048 292 26 — — PUNCT bracis-19048 292 27 the the DET bracis-19048 292 28 benefits benefit NOUN bracis-19048 292 29 of of ADP bracis-19048 292 30 multiple multiple ADJ bracis-19048 292 31 objectives objective NOUN bracis-19048 292 32 in in ADP bracis-19048 292 33 data datum NOUN bracis-19048 292 34 clustering cluster VERB bracis-19048 292 35 . . PUNCT bracis-19048 293 1 in in ADP bracis-19048 293 2 : : PUNCT bracis-19048 293 3 coello coello PROPN bracis-19048 293 4 coello coello PROPN bracis-19048 293 5 , , PUNCT bracis-19048 293 6 c.a c.a PROPN bracis-19048 293 7 . . PROPN bracis-19048 293 8 , , PUNCT bracis-19048 293 9 hernández hernández PROPN bracis-19048 293 10 aguirre aguirre PROPN bracis-19048 293 11 , , PUNCT bracis-19048 293 12 a. a. PROPN bracis-19048 293 13 , , PUNCT bracis-19048 293 14 zitzler zitzler NOUN bracis-19048 293 15 , , PUNCT bracis-19048 293 16 e. e. PROPN bracis-19048 293 17 ( ( PUNCT bracis-19048 293 18 eds eds PROPN bracis-19048 293 19 . . PUNCT bracis-19048 293 20 ) ) PUNCT bracis-19048 293 21 emo emo NOUN bracis-19048 293 22 2005 2005 NUM bracis-19048 293 23 . . PUNCT bracis-19048 294 1 lncs lncs PROPN bracis-19048 294 2 , , PUNCT bracis-19048 294 3 vol vol NOUN bracis-19048 294 4 . . PUNCT bracis-19048 294 5 3410 3410 NUM bracis-19048 294 6 , , PUNCT bracis-19048 294 7 pp pp ADP bracis-19048 294 8 . . PUNCT bracis-19048 295 1 547–560 547–560 NUM bracis-19048 295 2 . . PUNCT bracis-19048 295 3 springer springer NOUN bracis-19048 295 4 , , PUNCT bracis-19048 295 5 heidelberg heidelberg PROPN bracis-19048 295 6 ( ( PUNCT bracis-19048 295 7 2005 2005 NUM bracis-19048 295 8 ) ) PUNCT bracis-19048 295 9 . . PUNCT bracis-19048 296 1 https://doi.org/10.1007/978-3-540-31880-4_38 https://doi.org/10.1007/978-3-540-31880-4_38 PROPN bracis-19048 296 2 chapter chapter NOUN bracis-19048 296 3     SPACE bracis-19048 296 4 google google PROPN bracis-19048 296 5 scholar scholar NOUN bracis-19048 296 6     SPACE bracis-19048 296 7 haykin haykin PROPN bracis-19048 296 8 , , PUNCT bracis-19048 296 9 s.s s.s PROPN bracis-19048 296 10 . . PROPN bracis-19048 296 11 : : PUNCT bracis-19048 297 1 neural neural ADJ bracis-19048 297 2 networks network NOUN bracis-19048 297 3 : : PUNCT bracis-19048 297 4 a a DET bracis-19048 297 5 comprehensive comprehensive ADJ bracis-19048 297 6 foundation foundation NOUN bracis-19048 297 7 . . PUNCT bracis-19048 298 1 prentice prentice PROPN bracis-19048 298 2 hall hall PROPN bracis-19048 298 3 , , PUNCT bracis-19048 298 4 hoboken hoboken PROPN bracis-19048 298 5 ( ( PUNCT bracis-19048 298 6 1999 1999 NUM bracis-19048 298 7 ) ) PUNCT bracis-19048 298 8 google google PROPN bracis-19048 298 9 scholar scholar NOUN bracis-19048 298 10     SPACE bracis-19048 298 11 hubert hubert PROPN bracis-19048 298 12 , , PUNCT bracis-19048 298 13 l. l. PROPN bracis-19048 298 14 , , PUNCT bracis-19048 298 15 schultz schultz PROPN bracis-19048 298 16 , , PUNCT bracis-19048 298 17 j. j. PROPN bracis-19048 298 18 : : PUNCT bracis-19048 298 19 quadratic quadratic ADJ bracis-19048 298 20 assignment assignment NOUN bracis-19048 298 21 as as ADP bracis-19048 298 22 a a DET bracis-19048 298 23 general general ADJ bracis-19048 298 24 data datum NOUN bracis-19048 298 25 analysis analysis NOUN bracis-19048 298 26 strategy strategy NOUN bracis-19048 298 27 . . PUNCT bracis-19048 299 1 br br PROPN bracis-19048 299 2 . . PUNCT bracis-19048 300 1 j. j. PROPN bracis-19048 300 2 math math PROPN bracis-19048 300 3 . . PUNCT bracis-19048 301 1 stat stat PROPN bracis-19048 301 2 . . PUNCT bracis-19048 302 1 psychol psychol NOUN bracis-19048 302 2 . . PUNCT bracis-19048 303 1 29(2 29(2 X bracis-19048 303 2 ) ) PUNCT bracis-19048 303 3 , , PUNCT bracis-19048 303 4 190–241 190–241 NUM bracis-19048 303 5 ( ( PUNCT bracis-19048 303 6 1976 1976 NUM bracis-19048 303 7 ) ) PUNCT bracis-19048 303 8 article article NOUN bracis-19048 303 9     SPACE bracis-19048 303 10 mathscinet mathscinet NOUN bracis-19048 303 11     SPACE bracis-19048 303 12 google google PROPN bracis-19048 303 13 scholar scholar NOUN bracis-19048 303 14     SPACE bracis-19048 303 15 mitchell mitchell PROPN bracis-19048 303 16 , , PUNCT bracis-19048 303 17 t.m t.m PROPN bracis-19048 303 18 . . PROPN bracis-19048 303 19 : : PUNCT bracis-19048 303 20 machine machine NOUN bracis-19048 303 21 learning learning NOUN bracis-19048 303 22 . . PUNCT bracis-19048 304 1 mcgraw mcgraw PROPN bracis-19048 304 2 hill hill PROPN bracis-19048 304 3 series series PROPN bracis-19048 304 4 in in ADP bracis-19048 304 5 computer computer NOUN bracis-19048 304 6 science science NOUN bracis-19048 304 7 . . PUNCT bracis-19048 305 1 mcgraw mcgraw PROPN bracis-19048 305 2 hill hill PROPN bracis-19048 305 3 , , PUNCT bracis-19048 305 4 new new PROPN bracis-19048 305 5 york york PROPN bracis-19048 305 6 ( ( PUNCT bracis-19048 305 7 1997 1997 NUM bracis-19048 305 8 ) ) PUNCT bracis-19048 305 9 google google PROPN bracis-19048 305 10 scholar scholar NOUN bracis-19048 305 11     SPACE bracis-19048 305 12 montgomery montgomery PROPN bracis-19048 305 13 , , PUNCT bracis-19048 305 14 d.c d.c PROPN bracis-19048 305 15 . . PUNCT bracis-19048 305 16 : : PUNCT bracis-19048 306 1 design design NOUN bracis-19048 306 2 and and CCONJ bracis-19048 306 3 analysis analysis NOUN bracis-19048 306 4 of of ADP bracis-19048 306 5 experiments experiment NOUN bracis-19048 306 6 , , PUNCT bracis-19048 306 7 5th 5th ADJ bracis-19048 306 8 edn edn NOUN bracis-19048 306 9 . . PUNCT bracis-19048 307 1 wiley wiley PROPN bracis-19048 307 2 , , PUNCT bracis-19048 307 3 hoboken hoboken PROPN bracis-19048 307 4 ( ( PUNCT bracis-19048 307 5 2000 2000 NUM bracis-19048 307 6 ) ) PUNCT bracis-19048 307 7 google google PROPN bracis-19048 307 8 scholar scholar NOUN bracis-19048 307 9     SPACE bracis-19048 307 10 muñoz muñoz PROPN bracis-19048 307 11 , , PUNCT bracis-19048 307 12 m.a m.a PROPN bracis-19048 307 13 . . PROPN bracis-19048 307 14 , , PUNCT bracis-19048 307 15 villanova villanova PROPN bracis-19048 307 16 , , PUNCT bracis-19048 307 17 l. l. PROPN bracis-19048 307 18 , , PUNCT bracis-19048 307 19 baatar baatar NOUN bracis-19048 307 20 , , PUNCT bracis-19048 307 21 d. d. PROPN bracis-19048 307 22 , , PUNCT bracis-19048 307 23 smith smith PROPN bracis-19048 307 24 - - PUNCT bracis-19048 307 25 miles mile NOUN bracis-19048 307 26 , , PUNCT bracis-19048 307 27 k. k. PROPN bracis-19048 307 28 : : PUNCT bracis-19048 307 29 instance instance NOUN bracis-19048 307 30 spaces space VERB bracis-19048 307 31 for for ADP bracis-19048 307 32 machine machine NOUN bracis-19048 307 33 learning learn VERB bracis-19048 307 34 classification classification NOUN bracis-19048 307 35 . . PUNCT bracis-19048 308 1 mach mach NOUN bracis-19048 308 2 . . PUNCT bracis-19048 309 1 learn learn VERB bracis-19048 309 2 . . PUNCT bracis-19048 310 1 107(1 107(1 NUM bracis-19048 310 2 ) ) PUNCT bracis-19048 310 3 , , PUNCT bracis-19048 310 4 109–147 109–147 NUM bracis-19048 310 5 ( ( PUNCT bracis-19048 310 6 2018 2018 NUM bracis-19048 310 7 ) ) PUNCT bracis-19048 310 8 . . PUNCT bracis-19048 311 1 https://doi.org/10.1007/s10994-017-5629-5 https://doi.org/10.1007/s10994-017-5629-5 PROPN bracis-19048 311 2 article article NOUN bracis-19048 311 3     SPACE bracis-19048 311 4 mathscinet mathscinet NOUN bracis-19048 311 5     SPACE bracis-19048 311 6 math math NOUN bracis-19048 311 7     SPACE bracis-19048 311 8 google google PROPN bracis-19048 311 9 scholar scholar NOUN bracis-19048 311 10     SPACE bracis-19048 311 11 pfahringer pfahringer NOUN bracis-19048 311 12 , , PUNCT bracis-19048 311 13 b. b. PROPN bracis-19048 311 14 , , PUNCT bracis-19048 311 15 bensusan bensusan PROPN bracis-19048 311 16 , , PUNCT bracis-19048 311 17 h. h. PROPN bracis-19048 311 18 , , PUNCT bracis-19048 311 19 giraud giraud NOUN bracis-19048 311 20 - - PUNCT bracis-19048 311 21 carrier carrier NOUN bracis-19048 311 22 , , PUNCT bracis-19048 311 23 c.g c.g PROPN bracis-19048 311 24 . . PROPN bracis-19048 311 25 : : PUNCT bracis-19048 312 1 meta meta VERB bracis-19048 312 2 - - PUNCT bracis-19048 312 3 learning learning NOUN bracis-19048 312 4 by by ADP bracis-19048 312 5 landmarking landmarke VERB bracis-19048 312 6 various various ADJ bracis-19048 312 7 learning learning NOUN bracis-19048 312 8 algorithms algorithm NOUN bracis-19048 312 9 . . PUNCT bracis-19048 313 1 in in ADP bracis-19048 313 2 : : PUNCT bracis-19048 313 3 17th 17th ADJ bracis-19048 313 4 international international ADJ bracis-19048 313 5 conference conference NOUN bracis-19048 313 6 on on ADP bracis-19048 313 7 machine machine NOUN bracis-19048 313 8 learning learning NOUN bracis-19048 313 9 ( ( PUNCT bracis-19048 313 10 icml icml PROPN bracis-19048 313 11 ) ) PUNCT bracis-19048 313 12 , , PUNCT bracis-19048 313 13 pp pp ADP bracis-19048 313 14 . . PUNCT bracis-19048 314 1 743–750 743–750 NUM bracis-19048 314 2 ( ( PUNCT bracis-19048 314 3 2000 2000 NUM bracis-19048 314 4 ) ) PUNCT bracis-19048 314 5 google google PROPN bracis-19048 314 6 scholar scholar NOUN bracis-19048 314 7     SPACE bracis-19048 314 8 pimentel pimentel PROPN bracis-19048 314 9 , , PUNCT bracis-19048 314 10 b.a b.a PROPN bracis-19048 314 11 . . PROPN bracis-19048 314 12 , , PUNCT bracis-19048 314 13 de de X bracis-19048 314 14 carvalho carvalho NOUN bracis-19048 314 15 , , PUNCT bracis-19048 314 16 a.c.p.l.f a.c.p.l.f PROPN bracis-19048 314 17 . . PUNCT bracis-19048 314 18 : : PUNCT bracis-19048 315 1 a a DET bracis-19048 315 2 new new ADJ bracis-19048 315 3 data data NOUN bracis-19048 315 4 characterization characterization NOUN bracis-19048 315 5 for for ADP bracis-19048 315 6 selecting select VERB bracis-19048 315 7 clustering clustering ADJ bracis-19048 315 8 algorithms algorithm NOUN bracis-19048 315 9 using use VERB bracis-19048 315 10 meta meta NOUN bracis-19048 315 11 - - PUNCT bracis-19048 315 12 learning learning NOUN bracis-19048 315 13 . . PUNCT bracis-19048 316 1 inf inf PROPN bracis-19048 316 2 . . PUNCT bracis-19048 317 1 sci sci PROPN bracis-19048 317 2 . . PROPN bracis-19048 317 3 477 477 NUM bracis-19048 317 4 , , PUNCT bracis-19048 317 5 203–219 203–219 NUM bracis-19048 317 6 ( ( PUNCT bracis-19048 317 7 2019 2019 NUM bracis-19048 317 8 ) ) PUNCT bracis-19048 317 9 google google PROPN bracis-19048 317 10 scholar scholar NOUN bracis-19048 317 11     SPACE bracis-19048 317 12 quinlan quinlan PROPN bracis-19048 317 13 , , PUNCT bracis-19048 317 14 j.r j.r PROPN bracis-19048 317 15 . . PROPN bracis-19048 317 16 : : PUNCT bracis-19048 317 17 induction induction NOUN bracis-19048 317 18 of of ADP bracis-19048 317 19 decision decision NOUN bracis-19048 317 20 trees tree NOUN bracis-19048 317 21 . . PUNCT bracis-19048 318 1 mach mach NOUN bracis-19048 318 2 . . PUNCT bracis-19048 319 1 learn learn VERB bracis-19048 319 2 . . PUNCT bracis-19048 319 3 1(1 1(1 NUM bracis-19048 319 4 ) ) PUNCT bracis-19048 319 5 , , PUNCT bracis-19048 319 6 81–106 81–106 PROPN bracis-19048 319 7 ( ( PUNCT bracis-19048 319 8 1986 1986 NUM bracis-19048 319 9 ) ) PUNCT bracis-19048 319 10 article article NOUN bracis-19048 319 11     SPACE bracis-19048 319 12 google google PROPN bracis-19048 319 13 scholar scholar NOUN bracis-19048 319 14     SPACE bracis-19048 319 15 ray ray NOUN bracis-19048 319 16 , , PUNCT bracis-19048 319 17 s. s. PROPN bracis-19048 319 18 , , PUNCT bracis-19048 319 19 turi turi PROPN bracis-19048 319 20 , , PUNCT bracis-19048 319 21 r.h r.h PROPN bracis-19048 319 22 . . PROPN bracis-19048 319 23 : : PUNCT bracis-19048 320 1 determination determination NOUN bracis-19048 320 2 of of ADP bracis-19048 320 3 number number NOUN bracis-19048 320 4 of of ADP bracis-19048 320 5 clusters cluster NOUN bracis-19048 320 6 in in ADP bracis-19048 320 7 k k PROPN bracis-19048 320 8 - - PUNCT bracis-19048 320 9 means mean VERB bracis-19048 320 10 clustering clustering NOUN bracis-19048 320 11 and and CCONJ bracis-19048 320 12 application application NOUN bracis-19048 320 13 in in ADP bracis-19048 320 14 colour colour ADJ bracis-19048 320 15 segmentation segmentation NOUN bracis-19048 320 16 . . PUNCT bracis-19048 321 1 in in ADP bracis-19048 321 2 : : PUNCT bracis-19048 321 3 4th 4th ADJ bracis-19048 321 4 international international ADJ bracis-19048 321 5 conference conference NOUN bracis-19048 321 6 on on ADP bracis-19048 321 7 advances advance NOUN bracis-19048 321 8 in in ADP bracis-19048 321 9 pattern pattern NOUN bracis-19048 321 10 recognition recognition NOUN bracis-19048 321 11 and and CCONJ bracis-19048 321 12 digital digital ADJ bracis-19048 321 13 techniques technique NOUN bracis-19048 321 14 ( ( PUNCT bracis-19048 321 15 icaprdt icaprdt NOUN bracis-19048 321 16 ) ) PUNCT bracis-19048 321 17 , , PUNCT bracis-19048 321 18 pp pp ADP bracis-19048 321 19 . . PUNCT bracis-19048 322 1 137–143 137–143 NUM bracis-19048 322 2 ( ( PUNCT bracis-19048 322 3 1999 1999 NUM bracis-19048 322 4 ) ) PUNCT bracis-19048 322 5 google google PROPN bracis-19048 322 6 scholar scholar NOUN bracis-19048 322 7     SPACE bracis-19048 322 8 reif reif NOUN bracis-19048 322 9 , , PUNCT bracis-19048 322 10 m. m. NOUN bracis-19048 322 11 , , PUNCT bracis-19048 322 12 shafait shafait NOUN bracis-19048 322 13 , , PUNCT bracis-19048 322 14 f. f. PROPN bracis-19048 322 15 , , PUNCT bracis-19048 322 16 goldstein goldstein PROPN bracis-19048 322 17 , , PUNCT bracis-19048 322 18 m. m. NOUN bracis-19048 322 19 , , PUNCT bracis-19048 322 20 breuel breuel NOUN bracis-19048 322 21 , , PUNCT bracis-19048 322 22 t. t. PROPN bracis-19048 322 23 , , PUNCT bracis-19048 322 24 dengel dengel PROPN bracis-19048 322 25 , , PUNCT bracis-19048 322 26 a. a. NOUN bracis-19048 322 27 : : PUNCT bracis-19048 322 28 automatic automatic ADJ bracis-19048 322 29 classifier classifier NOUN bracis-19048 322 30 selection selection NOUN bracis-19048 322 31 for for ADP bracis-19048 322 32 non non NOUN bracis-19048 322 33 - - NOUN bracis-19048 322 34 experts expert NOUN bracis-19048 322 35 . . PUNCT bracis-19048 323 1 pattern pattern NOUN bracis-19048 323 2 anal anal PROPN bracis-19048 323 3 . . PUNCT bracis-19048 324 1 appl appl PROPN bracis-19048 324 2 . . PUNCT bracis-19048 325 1 17(1 17(1 NUM bracis-19048 325 2 ) ) PUNCT bracis-19048 325 3 , , PUNCT bracis-19048 325 4 83–96 83–96 NUM bracis-19048 325 5 ( ( PUNCT bracis-19048 325 6 2014 2014 NUM bracis-19048 325 7 ) ) PUNCT bracis-19048 325 8 article article NOUN bracis-19048 325 9     SPACE bracis-19048 325 10 mathscinet mathscinet NOUN bracis-19048 325 11     SPACE bracis-19048 325 12 google google PROPN bracis-19048 325 13 scholar scholar NOUN bracis-19048 325 14     SPACE bracis-19048 325 15 rice rice NOUN bracis-19048 325 16 , , PUNCT bracis-19048 325 17 j.r j.r PROPN bracis-19048 325 18 . . PROPN bracis-19048 325 19 : : PUNCT bracis-19048 326 1 the the DET bracis-19048 326 2 algorithm algorithm NOUN bracis-19048 326 3 selection selection NOUN bracis-19048 326 4 problem problem NOUN bracis-19048 326 5 . . PUNCT bracis-19048 327 1 adv adv INTJ bracis-19048 327 2 . . PUNCT bracis-19048 328 1 comput comput PROPN bracis-19048 328 2 . . PUNCT bracis-19048 329 1 15 15 NUM bracis-19048 329 2 , , PUNCT bracis-19048 329 3 65–118 65–118 NUM bracis-19048 329 4 ( ( PUNCT bracis-19048 329 5 1976 1976 NUM bracis-19048 329 6 ) ) PUNCT bracis-19048 329 7 article article NOUN bracis-19048 329 8     SPACE bracis-19048 329 9 google google PROPN bracis-19048 329 10 scholar scholar NOUN bracis-19048 329 11     SPACE bracis-19048 329 12 rivolli rivolli NOUN bracis-19048 329 13 , , PUNCT bracis-19048 329 14 a. a. NOUN bracis-19048 329 15 , , PUNCT bracis-19048 329 16 garcia garcia PROPN bracis-19048 329 17 , , PUNCT bracis-19048 329 18 l.p.f l.p.f VERB bracis-19048 329 19 . . PROPN bracis-19048 329 20 , , PUNCT bracis-19048 329 21 soares soares PROPN bracis-19048 329 22 , , PUNCT bracis-19048 329 23 c. c. PROPN bracis-19048 329 24 , , PUNCT bracis-19048 329 25 vanschoren vanschoren PROPN bracis-19048 329 26 , , PUNCT bracis-19048 329 27 j. j. PROPN bracis-19048 329 28 , , PUNCT bracis-19048 329 29 de de PROPN bracis-19048 329 30 carvalho carvalho NOUN bracis-19048 329 31 , , PUNCT bracis-19048 329 32 a.c.p.l.f a.c.p.l.f PROPN bracis-19048 329 33 . . PUNCT bracis-19048 329 34 : : PUNCT bracis-19048 330 1 characterizing characterize VERB bracis-19048 330 2 classification classification NOUN bracis-19048 330 3 datasets dataset NOUN bracis-19048 330 4 : : PUNCT bracis-19048 330 5 a a DET bracis-19048 330 6 study study NOUN bracis-19048 330 7 of of ADP bracis-19048 330 8 meta meta NOUN bracis-19048 330 9 - - PUNCT bracis-19048 330 10 features feature NOUN bracis-19048 330 11 for for ADP bracis-19048 330 12 meta meta NOUN bracis-19048 330 13 - - PUNCT bracis-19048 330 14 learning learning NOUN bracis-19048 330 15 . . PUNCT bracis-19048 331 1 corr corr PROPN bracis-19048 331 2 abs/1808.10406 abs/1808.10406 VERB bracis-19048 331 3 , , PUNCT bracis-19048 331 4 1–49 1–49 PROPN bracis-19048 331 5 ( ( PUNCT bracis-19048 331 6 2019 2019 NUM bracis-19048 331 7 ) ) PUNCT bracis-19048 331 8 google google NOUN bracis-19048 331 9 scholar scholar NOUN bracis-19048 331 10     SPACE bracis-19048 331 11 rousseeuw rousseeuw NOUN bracis-19048 331 12 , , PUNCT bracis-19048 331 13 p.j p.j PROPN bracis-19048 331 14 . . PROPN bracis-19048 331 15 : : PUNCT bracis-19048 332 1 silhouettes silhouette NOUN bracis-19048 332 2 : : PUNCT bracis-19048 332 3 a a DET bracis-19048 332 4 graphical graphical ADJ bracis-19048 332 5 aid aid NOUN bracis-19048 332 6 to to ADP bracis-19048 332 7 the the DET bracis-19048 332 8 interpretation interpretation NOUN bracis-19048 332 9 and and CCONJ bracis-19048 332 10 validation validation NOUN bracis-19048 332 11 of of ADP bracis-19048 332 12 cluster cluster NOUN bracis-19048 332 13 analysis analysis NOUN bracis-19048 332 14 . . PUNCT bracis-19048 333 1 j. j. PROPN bracis-19048 333 2 comput comput PROPN bracis-19048 333 3 . . PUNCT bracis-19048 334 1 appl appl PROPN bracis-19048 334 2 . . PROPN bracis-19048 334 3 math math NOUN bracis-19048 334 4 . . PUNCT bracis-19048 335 1 20 20 NUM bracis-19048 335 2 , , PUNCT bracis-19048 335 3 53–65 53–65 NUM bracis-19048 335 4 ( ( PUNCT bracis-19048 335 5 1987 1987 NUM bracis-19048 335 6 ) ) PUNCT bracis-19048 335 7 article article NOUN bracis-19048 335 8     SPACE bracis-19048 335 9 google google PROPN bracis-19048 335 10 scholar scholar NOUN bracis-19048 335 11     SPACE bracis-19048 335 12 sakamoto sakamoto NOUN bracis-19048 335 13 , , PUNCT bracis-19048 335 14 y. y. PROPN bracis-19048 335 15 , , PUNCT bracis-19048 335 16 ishiguro ishiguro PROPN bracis-19048 335 17 , , PUNCT bracis-19048 335 18 m. m. NOUN bracis-19048 335 19 , , PUNCT bracis-19048 335 20 kitagawa kitagawa PROPN bracis-19048 335 21 , , PUNCT bracis-19048 335 22 g. g. PROPN bracis-19048 335 23 : : PUNCT bracis-19048 335 24 akaike akaike ADJ bracis-19048 335 25 information information NOUN bracis-19048 335 26 criterion criterion NOUN bracis-19048 335 27 statistics statistic NOUN bracis-19048 335 28 . . PUNCT bracis-19048 336 1 springer springer PROPN bracis-19048 336 2 , , PUNCT bracis-19048 336 3 netherlands netherlands PROPN bracis-19048 336 4 ( ( PUNCT bracis-19048 336 5 1986 1986 NUM bracis-19048 336 6 ) ) PUNCT bracis-19048 336 7 math math NOUN bracis-19048 336 8     SPACE bracis-19048 336 9 google google PROPN bracis-19048 336 10 scholar scholar NOUN bracis-19048 336 11     SPACE bracis-19048 336 12 smith smith PROPN bracis-19048 336 13 - - PUNCT bracis-19048 336 14 miles miles PROPN bracis-19048 336 15 , , PUNCT bracis-19048 336 16 k.a k.a PROPN bracis-19048 336 17 . . PROPN bracis-19048 336 18 : : PUNCT bracis-19048 337 1 cross cross ADJ bracis-19048 337 2 - - ADJ bracis-19048 337 3 disciplinary disciplinary ADJ bracis-19048 337 4 perspectives perspective NOUN bracis-19048 337 5 on on ADP bracis-19048 337 6 meta meta NOUN bracis-19048 337 7 - - PUNCT bracis-19048 337 8 learning learning NOUN bracis-19048 337 9 for for ADP bracis-19048 337 10 algorithm algorithm NOUN bracis-19048 337 11 selection selection NOUN bracis-19048 337 12 . . PUNCT bracis-19048 338 1 acm acm PROPN bracis-19048 338 2 comput comput NOUN bracis-19048 338 3 . . PUNCT bracis-19048 339 1 surv surv PROPN bracis-19048 339 2 . . PUNCT bracis-19048 340 1 41(1 41(1 NUM bracis-19048 340 2 ) ) PUNCT bracis-19048 340 3 , , PUNCT bracis-19048 340 4 1–25 1–25 PROPN bracis-19048 340 5 ( ( PUNCT bracis-19048 340 6 2008 2008 NUM bracis-19048 340 7 ) ) PUNCT bracis-19048 340 8 article article NOUN bracis-19048 340 9     SPACE bracis-19048 340 10 google google PROPN bracis-19048 340 11 scholar scholar NOUN bracis-19048 340 12     SPACE bracis-19048 340 13 stephenson stephenson NOUN bracis-19048 340 14 , , PUNCT bracis-19048 340 15 n. n. PROPN bracis-19048 340 16 , , PUNCT bracis-19048 340 17 et et PROPN bracis-19048 340 18 al al PROPN bracis-19048 340 19 . . PUNCT bracis-19048 340 20 : : PUNCT bracis-19048 341 1 survey survey NOUN bracis-19048 341 2 of of ADP bracis-19048 341 3 machine machine NOUN bracis-19048 341 4 learning learn VERB bracis-19048 341 5 techniques technique NOUN bracis-19048 341 6 in in ADP bracis-19048 341 7 drug drug NOUN bracis-19048 341 8 discovery discovery NOUN bracis-19048 341 9 . . PUNCT bracis-19048 342 1 curr curr PROPN bracis-19048 342 2 . . PUNCT bracis-19048 342 3 drug drug PROPN bracis-19048 342 4 metab metab NOUN bracis-19048 342 5 . . PUNCT bracis-19048 343 1 20(3 20(3 NOUN bracis-19048 343 2 ) ) PUNCT bracis-19048 343 3 , , PUNCT bracis-19048 343 4 185–193 185–193 NUM bracis-19048 343 5 ( ( PUNCT bracis-19048 343 6 2019 2019 NUM bracis-19048 343 7 ) ) PUNCT bracis-19048 343 8 article article NOUN bracis-19048 343 9     SPACE bracis-19048 343 10 google google PROPN bracis-19048 343 11 scholar scholar NOUN bracis-19048 343 12     SPACE bracis-19048 343 13 van van PROPN bracis-19048 343 14 rijn rijn PROPN bracis-19048 343 15 , , PUNCT bracis-19048 343 16 j.n j.n PROPN bracis-19048 343 17 . . PROPN bracis-19048 343 18 , , PUNCT bracis-19048 343 19 et et PROPN bracis-19048 343 20 al al PROPN bracis-19048 343 21 . . PROPN bracis-19048 343 22 : : PUNCT bracis-19048 344 1 openml openml NOUN bracis-19048 344 2 : : PUNCT bracis-19048 344 3 a a DET bracis-19048 344 4 collaborative collaborative ADJ bracis-19048 344 5 science science NOUN bracis-19048 344 6 platform platform NOUN bracis-19048 344 7 . . PUNCT bracis-19048 345 1 in in ADP bracis-19048 345 2 : : PUNCT bracis-19048 345 3 european european ADJ bracis-19048 345 4 conference conference NOUN bracis-19048 345 5 on on ADP bracis-19048 345 6 machine machine NOUN bracis-19048 345 7 learning learning NOUN bracis-19048 345 8 and and CCONJ bracis-19048 345 9 knowledge knowledge NOUN bracis-19048 345 10 discovery discovery NOUN bracis-19048 345 11 in in ADP bracis-19048 345 12 databases database NOUN bracis-19048 345 13 ( ( PUNCT bracis-19048 345 14 ecml ecml NOUN bracis-19048 345 15 / / SYM bracis-19048 345 16 pkdd pkdd NOUN bracis-19048 345 17 ) ) PUNCT bracis-19048 345 18 , , PUNCT bracis-19048 345 19 pp pp ADJ bracis-19048 345 20 . . PUNCT bracis-19048 346 1 645–649 645–649 NUM bracis-19048 346 2 ( ( PUNCT bracis-19048 346 3 2013 2013 NUM bracis-19048 346 4 ) ) PUNCT bracis-19048 346 5 google google PROPN bracis-19048 346 6 scholar scholar NOUN bracis-19048 346 7     SPACE bracis-19048 346 8 vukicevic vukicevic NOUN bracis-19048 346 9 , , PUNCT bracis-19048 346 10 m. m. NOUN bracis-19048 346 11 , , PUNCT bracis-19048 346 12 radovanovic radovanovic VERB bracis-19048 346 13 , , PUNCT bracis-19048 346 14 s. s. PROPN bracis-19048 346 15 , , PUNCT bracis-19048 346 16 delibasic delibasic PROPN bracis-19048 346 17 , , PUNCT bracis-19048 346 18 b. b. PROPN bracis-19048 346 19 , , PUNCT bracis-19048 346 20 suknovic suknovic ADJ bracis-19048 346 21 , , PUNCT bracis-19048 346 22 m. m. NOUN bracis-19048 346 23 : : PUNCT bracis-19048 346 24 extending extend VERB bracis-19048 346 25 meta meta ADJ bracis-19048 346 26 - - PUNCT bracis-19048 346 27 learning learn VERB bracis-19048 346 28 framework framework NOUN bracis-19048 346 29 for for ADP bracis-19048 346 30 clustering cluster VERB bracis-19048 346 31 gene gene NOUN bracis-19048 346 32 expression expression NOUN bracis-19048 346 33 data datum NOUN bracis-19048 346 34 with with ADP bracis-19048 346 35 component component NOUN bracis-19048 346 36 - - PUNCT bracis-19048 346 37 based base VERB bracis-19048 346 38 algorithm algorithm NOUN bracis-19048 346 39 design design NOUN bracis-19048 346 40 and and CCONJ bracis-19048 346 41 internal internal ADJ bracis-19048 346 42 evaluation evaluation NOUN bracis-19048 346 43 measures measure NOUN bracis-19048 346 44 . . PUNCT bracis-19048 347 1 int int NOUN bracis-19048 347 2 . . PUNCT bracis-19048 348 1 j. j. PROPN bracis-19048 348 2 data data PROPN bracis-19048 348 3 min min PROPN bracis-19048 348 4 . . PROPN bracis-19048 348 5 bioinform bioinform PROPN bracis-19048 348 6 . . PUNCT bracis-19048 349 1 ( ( PUNCT bracis-19048 349 2 ijdmb ijdmb PROPN bracis-19048 349 3 ) ) PUNCT bracis-19048 349 4 14(2 14(2 NUM bracis-19048 349 5 ) ) PUNCT bracis-19048 349 6 , , PUNCT bracis-19048 349 7 101–119 101–119 NUM bracis-19048 349 8 ( ( PUNCT bracis-19048 349 9 2016 2016 NUM bracis-19048 349 10 ) ) PUNCT bracis-19048 349 11 article article NOUN bracis-19048 349 12     SPACE bracis-19048 349 13 google google PROPN bracis-19048 349 14 scholar scholar NOUN bracis-19048 349 15     SPACE bracis-19048 349 16 xie xie PROPN bracis-19048 349 17 , , PUNCT bracis-19048 349 18 x.l x.l PROPN bracis-19048 349 19 . . PROPN bracis-19048 349 20 , , PUNCT bracis-19048 349 21 beni beni NOUN bracis-19048 349 22 , , PUNCT bracis-19048 349 23 g. g. PROPN bracis-19048 349 24 : : PUNCT bracis-19048 349 25 a a DET bracis-19048 349 26 validity validity NOUN bracis-19048 349 27 measure measure NOUN bracis-19048 349 28 for for ADP bracis-19048 349 29 fuzzy fuzzy ADJ bracis-19048 349 30 clustering clustering NOUN bracis-19048 349 31 . . PUNCT bracis-19048 350 1 ieee ieee PROPN bracis-19048 350 2 trans trans PROPN bracis-19048 350 3 . . PUNCT bracis-19048 350 4 pattern pattern PROPN bracis-19048 350 5 anal anal PROPN bracis-19048 350 6 . . PUNCT bracis-19048 351 1 mach mach PROPN bracis-19048 351 2 . . PUNCT bracis-19048 352 1 intell intell PROPN bracis-19048 352 2 . . PUNCT bracis-19048 353 1 13(8 13(8 NUM bracis-19048 353 2 ) ) PUNCT bracis-19048 353 3 , , PUNCT bracis-19048 353 4 841–847 841–847 PROPN bracis-19048 353 5 ( ( PUNCT bracis-19048 353 6 1991 1991 NUM bracis-19048 353 7 ) ) PUNCT bracis-19048 353 8 article article NOUN bracis-19048 353 9     SPACE bracis-19048 353 10 google google PROPN bracis-19048 353 11 scholar scholar NOUN bracis-19048 353 12     SPACE bracis-19048 353 13 download download NOUN bracis-19048 353 14 references reference NOUN bracis-19048 353 15 acknowledgment acknowledgment NOUN bracis-19048 353 16 research research NOUN bracis-19048 353 17 carried carry VERB bracis-19048 353 18 out out ADP bracis-19048 353 19 using use VERB bracis-19048 353 20 the the DET bracis-19048 353 21 computational computational ADJ bracis-19048 353 22 resources resource NOUN bracis-19048 353 23 of of ADP bracis-19048 353 24 the the DET bracis-19048 353 25 center center NOUN bracis-19048 353 26 for for ADP bracis-19048 353 27 mathematical mathematical ADJ bracis-19048 353 28 sciences science NOUN bracis-19048 353 29 applied apply VERB bracis-19048 353 30 to to ADP bracis-19048 353 31 industry industry NOUN bracis-19048 353 32 ( ( PUNCT bracis-19048 353 33 cemeai cemeai PROPN bracis-19048 353 34 ) ) PUNCT bracis-19048 353 35 funded fund VERB bracis-19048 353 36 by by ADP bracis-19048 353 37 fapesp fapesp NOUN bracis-19048 353 38 ( ( PUNCT bracis-19048 353 39 grant grant NOUN bracis-19048 353 40 2013/07375 2013/07375 NUM bracis-19048 353 41 - - SYM bracis-19048 353 42 0 0 NUM bracis-19048 353 43 ) ) PUNCT bracis-19048 353 44 . . PUNCT bracis-19048 354 1 author author NOUN bracis-19048 354 2 information information NOUN bracis-19048 354 3 authors author NOUN bracis-19048 354 4 and and CCONJ bracis-19048 354 5 affiliations affiliation NOUN bracis-19048 354 6 department department PROPN bracis-19048 354 7 of of ADP bracis-19048 354 8 computer computer NOUN bracis-19048 354 9 science science NOUN bracis-19048 354 10 , , PUNCT bracis-19048 354 11 university university NOUN bracis-19048 354 12 of of ADP bracis-19048 354 13 brasília brasília PROPN bracis-19048 354 14 , , PUNCT bracis-19048 354 15 brasília brasília PROPN bracis-19048 354 16 , , PUNCT bracis-19048 354 17 brazil brazil PROPN bracis-19048 354 18 luís luís PROPN bracis-19048 355 1 p. p. PROPN bracis-19048 355 2 f. f. PROPN bracis-19048 355 3 garcia garcia PROPN bracis-19048 355 4     SPACE bracis-19048 355 5 & & CCONJ bracis-19048 355 6     SPACE bracis-19048 355 7 guilherme guilherme PROPN bracis-19048 355 8 n. n. PROPN bracis-19048 355 9 ramos ramos PROPN bracis-19048 355 10 college college PROPN bracis-19048 355 11 of of ADP bracis-19048 355 12 engineering engineering NOUN bracis-19048 355 13 and and CCONJ bracis-19048 355 14 physical physical ADJ bracis-19048 355 15 sciences sciences PROPN bracis-19048 355 16 , , PUNCT bracis-19048 355 17 aston aston PROPN bracis-19048 355 18 university university PROPN bracis-19048 355 19 , , PUNCT bracis-19048 355 20 birmingham birmingham PROPN bracis-19048 355 21 , , PUNCT bracis-19048 355 22 uk uk PROPN bracis-19048 355 23 felipe felipe PROPN bracis-19048 355 24 campelo campelo PROPN bracis-19048 355 25 computing computing PROPN bracis-19048 355 26 department department PROPN bracis-19048 355 27 , , PUNCT bracis-19048 355 28 technological technological ADJ bracis-19048 355 29 university university NOUN bracis-19048 355 30 of of ADP bracis-19048 355 31 paraná paraná PROPN bracis-19048 355 32 , , PUNCT bracis-19048 355 33 cornélio cornélio NOUN bracis-19048 355 34 procópio procópio PROPN bracis-19048 355 35 , , PUNCT bracis-19048 355 36 brazil brazil PROPN bracis-19048 355 37 adriano adriano PROPN bracis-19048 355 38 rivolli rivolli PROPN bracis-19048 355 39 institute institute PROPN bracis-19048 355 40 of of ADP bracis-19048 355 41 mathematical mathematical ADJ bracis-19048 355 42 and and CCONJ bracis-19048 355 43 computer computer NOUN bracis-19048 355 44 sciences science NOUN bracis-19048 355 45 , , PUNCT bracis-19048 355 46 university university NOUN bracis-19048 355 47 of of ADP bracis-19048 355 48 são são PROPN bracis-19048 355 49 paulo paulo PROPN bracis-19048 355 50 , , PUNCT bracis-19048 355 51 são são PROPN bracis-19048 355 52 carlos carlos PROPN bracis-19048 355 53 , , PUNCT bracis-19048 355 54 brazil brazil PROPN bracis-19048 355 55 andré andré VERB bracis-19048 355 56 c. c. PROPN bracis-19048 355 57 p. p. PROPN bracis-19048 355 58 de de PROPN bracis-19048 355 59 l. l. PROPN bracis-19048 355 60 f. f. PROPN bracis-19048 355 61 de de PROPN bracis-19048 355 62 carvalho carvalho PROPN bracis-19048 355 63 authors author NOUN bracis-19048 356 1 luís luís PROPN bracis-19048 356 2 p. p. PROPN bracis-19048 356 3 f. f. PROPN bracis-19048 356 4 garciaview garciaview PROPN bracis-19048 356 5 author author NOUN bracis-19048 356 6 publications publication NOUN bracis-19048 356 7 search search NOUN bracis-19048 356 8 author author NOUN bracis-19048 356 9 on on ADP bracis-19048 356 10 : : PUNCT bracis-19048 356 11 pubmed pubmed PROPN bracis-19048 356 12     SPACE bracis-19048 356 13 google google PROPN bracis-19048 356 14 scholar scholar NOUN bracis-19048 356 15 felipe felipe PROPN bracis-19048 356 16 campeloview campeloview ADJ bracis-19048 356 17 author author NOUN bracis-19048 356 18 publications publication NOUN bracis-19048 356 19 search search NOUN bracis-19048 356 20 author author NOUN bracis-19048 356 21 on on ADP bracis-19048 356 22 : : PUNCT bracis-19048 356 23 pubmed pubmed PROPN bracis-19048 356 24     SPACE bracis-19048 356 25 google google PROPN bracis-19048 356 26 scholar scholar NOUN bracis-19048 356 27 guilherme guilherme PROPN bracis-19048 356 28 n. n. PROPN bracis-19048 356 29 ramosview ramosview PROPN bracis-19048 356 30 author author NOUN bracis-19048 356 31 publications publication VERB bracis-19048 356 32 search search NOUN bracis-19048 356 33 author author NOUN bracis-19048 356 34 on on ADP bracis-19048 356 35 : : PUNCT bracis-19048 356 36 pubmed pubmed PROPN bracis-19048 356 37     SPACE bracis-19048 356 38 google google PROPN bracis-19048 356 39 scholar scholar NOUN bracis-19048 356 40 adriano adriano PROPN bracis-19048 356 41 rivolliview rivolliview NOUN bracis-19048 356 42 author author NOUN bracis-19048 356 43 publications publication NOUN bracis-19048 356 44 search search NOUN bracis-19048 356 45 author author NOUN bracis-19048 356 46 on on ADP bracis-19048 356 47 : : PUNCT bracis-19048 356 48 pubmed pubmed PROPN bracis-19048 356 49     SPACE bracis-19048 356 50 google google PROPN bracis-19048 356 51 scholar scholar NOUN bracis-19048 356 52 andré andré PROPN bracis-19048 356 53 c. c. PROPN bracis-19048 356 54 p. p. PROPN bracis-19048 356 55 de de PROPN bracis-19048 356 56 l. l. PROPN bracis-19048 356 57 f. f. PROPN bracis-19048 356 58 de de PROPN bracis-19048 356 59 carvalhoview carvalhoview PROPN bracis-19048 356 60 author author NOUN bracis-19048 356 61 publications publication NOUN bracis-19048 356 62 search search NOUN bracis-19048 356 63 author author NOUN bracis-19048 356 64 on on ADP bracis-19048 356 65 : : PUNCT bracis-19048 356 66 pubmed pubmed PROPN bracis-19048 356 67     SPACE bracis-19048 356 68 google google PROPN bracis-19048 356 69 scholar scholar NOUN bracis-19048 356 70 corresponding correspond VERB bracis-19048 356 71 author author NOUN bracis-19048 356 72 correspondence correspondence NOUN bracis-19048 356 73 to to ADP bracis-19048 356 74 luís luís PROPN bracis-19048 357 1 p. p. PROPN bracis-19048 357 2 f. f. PROPN bracis-19048 357 3 garcia garcia PROPN bracis-19048 357 4 . . PUNCT bracis-19048 358 1 editor editor NOUN bracis-19048 358 2 information information NOUN bracis-19048 358 3 editors editor NOUN bracis-19048 358 4 and and CCONJ bracis-19048 358 5 affiliations affiliation NOUN bracis-19048 358 6 universidade universidade PROPN bracis-19048 358 7 federal federal PROPN bracis-19048 358 8 de de X bracis-19048 358 9 sergipe sergipe PROPN bracis-19048 358 10 , , PUNCT bracis-19048 358 11 são são NOUN bracis-19048 358 12 cristóvão cristóvão PROPN bracis-19048 358 13 , , PUNCT bracis-19048 358 14 brazil brazil PROPN bracis-19048 358 15 andré andré PROPN bracis-19048 358 16 britto britto PROPN bracis-19048 358 17 universidade universidade PROPN bracis-19048 358 18 de de PROPN bracis-19048 358 19 são são PROPN bracis-19048 358 20 paulo paulo PROPN bracis-19048 358 21 , , PUNCT bracis-19048 358 22 são são PROPN bracis-19048 358 23 paulo paulo PROPN bracis-19048 358 24 , , PUNCT bracis-19048 358 25 brazil brazil PROPN bracis-19048 358 26 karina karina PROPN bracis-19048 358 27 valdivia valdivia PROPN bracis-19048 358 28 delgado delgado VERB bracis-19048 358 29 rights right NOUN bracis-19048 358 30 and and CCONJ bracis-19048 358 31 permissions permission NOUN bracis-19048 358 32 reprints reprint NOUN bracis-19048 358 33 and and CCONJ bracis-19048 358 34 permissions permission VERB bracis-19048 358 35 copyright copyright NOUN bracis-19048 358 36 information information NOUN bracis-19048 358 37 © © PROPN bracis-19048 358 38 2021 2021 NUM bracis-19048 358 39 springer springer NOUN bracis-19048 358 40 nature nature NOUN bracis-19048 358 41 switzerland switzerland PROPN bracis-19048 358 42 ag ag PROPN bracis-19048 358 43 about about ADP bracis-19048 358 44 this this DET bracis-19048 358 45 paper paper NOUN bracis-19048 358 46 cite cite VERB bracis-19048 358 47 this this DET bracis-19048 358 48 paper paper NOUN bracis-19048 358 49 garcia garcia PROPN bracis-19048 358 50 , , PUNCT bracis-19048 358 51 l.p.f l.p.f VERB bracis-19048 358 52 . . PROPN bracis-19048 358 53 , , PUNCT bracis-19048 358 54 campelo campelo PROPN bracis-19048 358 55 , , PUNCT bracis-19048 358 56 f. f. PROPN bracis-19048 358 57 , , PUNCT bracis-19048 358 58 ramos ramos PROPN bracis-19048 358 59 , , PUNCT bracis-19048 358 60 g.n g.n PROPN bracis-19048 358 61 . . PROPN bracis-19048 358 62 , , PUNCT bracis-19048 358 63 rivolli rivolli PROPN bracis-19048 358 64 , , PUNCT bracis-19048 358 65 a. a. NOUN bracis-19048 358 66 , , PUNCT bracis-19048 358 67 de de X bracis-19048 358 68 carvalho carvalho NOUN bracis-19048 358 69 , , PUNCT bracis-19048 358 70 a.c.p.d.l.f a.c.p.d.l.f PROPN bracis-19048 358 71 . . PROPN bracis-19048 358 72 ( ( PUNCT bracis-19048 358 73 2021 2021 NUM bracis-19048 358 74 ) ) PUNCT bracis-19048 358 75 . . PUNCT bracis-19048 359 1 evaluating evaluate VERB bracis-19048 359 2 clustering cluster VERB bracis-19048 359 3 meta meta NOUN bracis-19048 359 4 - - PUNCT bracis-19048 359 5 features feature NOUN bracis-19048 359 6 for for ADP bracis-19048 359 7 classifier classifier ADJ bracis-19048 359 8 recommendation recommendation NOUN bracis-19048 359 9 . . PUNCT bracis-19048 360 1 in in ADP bracis-19048 360 2 : : PUNCT bracis-19048 360 3 britto britto PROPN bracis-19048 360 4 , , PUNCT bracis-19048 360 5 a. a. PROPN bracis-19048 360 6 , , PUNCT bracis-19048 360 7 valdivia valdivia PROPN bracis-19048 360 8 delgado delgado PROPN bracis-19048 360 9 , , PUNCT bracis-19048 360 10 k. k. PROPN bracis-19048 360 11 ( ( PUNCT bracis-19048 360 12 eds eds PROPN bracis-19048 360 13 ) ) PUNCT bracis-19048 360 14 intelligent intelligent ADJ bracis-19048 360 15 systems system NOUN bracis-19048 360 16 . . PUNCT bracis-19048 361 1 bracis bracis PROPN bracis-19048 361 2 2021 2021 NUM bracis-19048 361 3 . . PUNCT bracis-19048 362 1 lecture lecture NOUN bracis-19048 362 2 notes note NOUN bracis-19048 362 3 in in ADP bracis-19048 362 4 computer computer NOUN bracis-19048 362 5 science science NOUN bracis-19048 362 6 ( ( PUNCT bracis-19048 362 7 ) ) PUNCT bracis-19048 362 8 , , PUNCT bracis-19048 362 9 vol vol NOUN bracis-19048 362 10 13073 13073 NUM bracis-19048 362 11 . . PUNCT bracis-19048 363 1 springer springer NOUN bracis-19048 363 2 , , PUNCT bracis-19048 363 3 cham cham PROPN bracis-19048 363 4 . . PUNCT bracis-19048 364 1 https://doi.org/10.1007/978-3-030-91702-9_30 https://doi.org/10.1007/978-3-030-91702-9_30 PROPN bracis-19048 364 2 download download NOUN bracis-19048 364 3 citation citation NOUN bracis-19048 364 4 .ris .ris PUNCT bracis-19048 364 5 .enw .enw PROPN bracis-19048 364 6 .bib .bib PUNCT bracis-19048 365 1 doi doi PROPN bracis-19048 365 2 : : PUNCT bracis-19048 365 3 https://doi.org/10.1007/978-3-030-91702-9_30 https://doi.org/10.1007/978-3-030-91702-9_30 PROPN bracis-19048 365 4 published publish VERB bracis-19048 365 5 : : PUNCT bracis-19048 365 6 28 28 NUM bracis-19048 365 7 november november PROPN bracis-19048 365 8 2021 2021 NUM bracis-19048 365 9 publisher publisher NOUN bracis-19048 365 10 name name NOUN bracis-19048 365 11 : : PUNCT bracis-19048 365 12 springer springer NOUN bracis-19048 365 13 , , PUNCT bracis-19048 365 14 cham cham PROPN bracis-19048 365 15 print print PROPN bracis-19048 365 16 isbn isbn PROPN bracis-19048 365 17 : : PUNCT bracis-19048 365 18 978 978 NUM bracis-19048 365 19 - - SYM bracis-19048 365 20 3 3 NUM bracis-19048 365 21 - - PUNCT bracis-19048 365 22 030 030 NUM bracis-19048 365 23 - - PUNCT bracis-19048 365 24 91701 91701 NUM bracis-19048 365 25 - - SYM bracis-19048 365 26 2 2 NUM bracis-19048 365 27 online online ADJ bracis-19048 365 28 isbn isbn NOUN bracis-19048 365 29 : : PUNCT bracis-19048 365 30 978 978 NUM bracis-19048 365 31 - - SYM bracis-19048 365 32 3 3 NUM bracis-19048 365 33 - - PUNCT bracis-19048 365 34 030 030 NUM bracis-19048 365 35 - - PUNCT bracis-19048 365 36 91702 91702 NUM bracis-19048 365 37 - - SYM bracis-19048 365 38 9 9 NUM bracis-19048 365 39 ebook ebook NOUN bracis-19048 365 40 packages package NOUN bracis-19048 365 41 : : PUNCT bracis-19048 365 42 computer computer NOUN bracis-19048 365 43 sciencecomputer sciencecomputer NOUN bracis-19048 365 44 science science NOUN bracis-19048 365 45 ( ( PUNCT bracis-19048 365 46 r0 r0 NOUN bracis-19048 365 47 ) ) PUNCT bracis-19048 365 48 share share VERB bracis-19048 365 49 this this DET bracis-19048 365 50 paper paper NOUN bracis-19048 365 51 anyone anyone PRON bracis-19048 365 52 you you PRON bracis-19048 365 53 share share VERB bracis-19048 365 54 the the DET bracis-19048 365 55 following follow VERB bracis-19048 365 56 link link NOUN bracis-19048 365 57 with with ADP bracis-19048 365 58 will will AUX bracis-19048 365 59 be be AUX bracis-19048 365 60 able able ADJ bracis-19048 365 61 to to PART bracis-19048 365 62 read read VERB bracis-19048 365 63 this this DET bracis-19048 365 64 content content NOUN bracis-19048 365 65 : : PUNCT bracis-19048 365 66 get get VERB bracis-19048 365 67 shareable shareable ADJ bracis-19048 365 68 linksorry linksorry NOUN bracis-19048 365 69 , , PUNCT bracis-19048 365 70 a a DET bracis-19048 365 71 shareable shareable ADJ bracis-19048 365 72 link link NOUN bracis-19048 365 73 is be AUX bracis-19048 365 74 not not PART bracis-19048 365 75 currently currently ADV bracis-19048 365 76 available available ADJ bracis-19048 365 77 for for ADP bracis-19048 365 78 this this DET bracis-19048 365 79 article article NOUN bracis-19048 365 80 . . PUNCT bracis-19048 366 1 copy copy VERB bracis-19048 366 2 shareable shareable ADJ bracis-19048 366 3 link link NOUN bracis-19048 366 4 to to PART bracis-19048 366 5 clipboard clipboard NOUN bracis-19048 366 6 provided provide VERB bracis-19048 366 7 by by ADP bracis-19048 366 8 the the DET bracis-19048 366 9 springer springer NOUN bracis-19048 366 10 nature nature PROPN bracis-19048 366 11 sharedit sharedit PROPN bracis-19048 366 12 content content NOUN bracis-19048 366 13 - - PUNCT bracis-19048 366 14 sharing share VERB bracis-19048 366 15 initiative initiative NOUN bracis-19048 366 16 keywords keyword VERB bracis-19048 366 17 meta meta VERB bracis-19048 366 18 - - PUNCT bracis-19048 366 19 learning learn VERB bracis-19048 366 20 meta meta VERB bracis-19048 366 21 - - PUNCT bracis-19048 366 22 features feature NOUN bracis-19048 366 23 characterization characterization NOUN bracis-19048 366 24 measures measure NOUN bracis-19048 366 25 clustering cluster VERB bracis-19048 366 26 problems problem NOUN bracis-19048 366 27 publish publish VERB bracis-19048 366 28 with with ADP bracis-19048 366 29 us us PROPN bracis-19048 366 30 policies policy NOUN bracis-19048 366 31 and and CCONJ bracis-19048 366 32 ethics ethic NOUN bracis-19048 366 33 profiles profile NOUN bracis-19048 366 34 felipe felipe PROPN bracis-19048 366 35 campelo campelo NOUN bracis-19048 366 36 view view NOUN bracis-19048 366 37 author author NOUN bracis-19048 366 38 profile profile PROPN bracis-19048 366 39 andré andré PROPN bracis-19048 366 40 c. c. PROPN bracis-19048 366 41 p. p. PROPN bracis-19048 366 42 de de PROPN bracis-19048 366 43 l. l. PROPN bracis-19048 366 44 f. f. PROPN bracis-19048 366 45 de de PROPN bracis-19048 366 46 carvalho carvalho PROPN bracis-19048 366 47 view view PROPN bracis-19048 366 48 author author NOUN bracis-19048 366 49 profile profile NOUN bracis-19048 366 50 search search NOUN bracis-19048 366 51 search search NOUN bracis-19048 366 52 by by ADP bracis-19048 366 53 keyword keyword NOUN bracis-19048 366 54 or or CCONJ bracis-19048 366 55 author author NOUN bracis-19048 366 56 search search NOUN bracis-19048 366 57 navigation navigation NOUN bracis-19048 366 58 find find VERB bracis-19048 366 59 a a DET bracis-19048 366 60 journal journal NOUN bracis-19048 366 61 publish publish VERB bracis-19048 366 62 with with ADP bracis-19048 366 63 us we PRON bracis-19048 366 64 track track VERB bracis-19048 366 65 your your PRON bracis-19048 366 66 research research NOUN bracis-19048 366 67 discover discover VERB bracis-19048 366 68 content content NOUN bracis-19048 366 69 journals journal NOUN bracis-19048 366 70 a a DET bracis-19048 366 71 - - PUNCT bracis-19048 366 72 z z NOUN bracis-19048 366 73 books book NOUN bracis-19048 366 74 a a DET bracis-19048 366 75 - - PUNCT bracis-19048 366 76 z z NOUN bracis-19048 366 77 publish publish NOUN bracis-19048 366 78 with with ADP bracis-19048 366 79 us us PROPN bracis-19048 366 80 journal journal PROPN bracis-19048 366 81 finder finder PROPN bracis-19048 366 82 publish publish VERB bracis-19048 366 83 your your PRON bracis-19048 366 84 research research NOUN bracis-19048 366 85 language language NOUN bracis-19048 366 86 editing edit VERB bracis-19048 366 87 open open ADJ bracis-19048 366 88 access access NOUN bracis-19048 366 89 publishing publishing NOUN bracis-19048 366 90 products product NOUN bracis-19048 366 91 and and CCONJ bracis-19048 366 92 services service NOUN bracis-19048 366 93 our our PRON bracis-19048 366 94 products product NOUN bracis-19048 366 95 librarians librarian VERB bracis-19048 366 96 societies society NOUN bracis-19048 366 97 partners partner NOUN bracis-19048 366 98 and and CCONJ bracis-19048 366 99 advertisers advertiser NOUN bracis-19048 366 100 our our PRON bracis-19048 366 101 brands brand NOUN bracis-19048 366 102 springer springer NOUN bracis-19048 366 103 nature nature PROPN bracis-19048 366 104 portfolio portfolio PROPN bracis-19048 366 105 bmc bmc PROPN bracis-19048 366 106 palgrave palgrave PROPN bracis-19048 366 107 macmillan macmillan PROPN bracis-19048 366 108 apress apress PROPN bracis-19048 366 109 discover discover VERB bracis-19048 366 110 your your PRON bracis-19048 366 111 privacy privacy NOUN bracis-19048 366 112 choices choice NOUN bracis-19048 366 113 / / SYM bracis-19048 366 114 manage manage NOUN bracis-19048 366 115 cookies cookie NOUN bracis-19048 366 116 your your PRON bracis-19048 366 117 us us PROPN bracis-19048 367 1 state state NOUN bracis-19048 367 2 privacy privacy NOUN bracis-19048 367 3 rights right NOUN bracis-19048 367 4 accessibility accessibility NOUN bracis-19048 367 5 statement statement NOUN bracis-19048 367 6 terms term NOUN bracis-19048 367 7 and and CCONJ bracis-19048 367 8 conditions condition NOUN bracis-19048 367 9 privacy privacy NOUN bracis-19048 367 10 policy policy NOUN bracis-19048 367 11 help help NOUN bracis-19048 367 12 and and CCONJ bracis-19048 367 13 support support VERB bracis-19048 367 14 legal legal ADJ bracis-19048 367 15 notice notice NOUN bracis-19048 367 16 cancel cancel VERB bracis-19048 367 17 contracts contract NOUN bracis-19048 367 18 here here ADV bracis-19048 367 19 129.74.145.123 129.74.145.123 NUM bracis-19048 367 20 hesburgh hesburgh PROPN bracis-19048 367 21 library library PROPN bracis-19048 367 22 er er INTJ bracis-19048 367 23 unit unit NOUN bracis-19048 367 24 ( ( PUNCT bracis-19048 367 25 3005732405 3005732405 NUM bracis-19048 367 26 ) ) PUNCT bracis-19048 367 27 northeast northeast ADJ bracis-19048 367 28 research research NOUN bracis-19048 367 29 libraries library NOUN bracis-19048 367 30 ( ( PUNCT bracis-19048 367 31 nerl nerl PROPN bracis-19048 367 32 ) ) PUNCT bracis-19048 367 33 ( ( PUNCT bracis-19048 367 34 8200828607 8200828607 NUM bracis-19048 367 35 ) ) PUNCT bracis-19048 367 36 nerl nerl VERB bracis-19048 367 37 ta ta X bracis-19048 367 38 account account NOUN bracis-19048 367 39 ( ( PUNCT bracis-19048 367 40 3006206169 3006206169 NUM bracis-19048 367 41 ) ) PUNCT bracis-19048 367 42 university university NOUN bracis-19048 367 43 of of ADP bracis-19048 367 44 notre notre PROPN bracis-19048 367 45 dame dame PROPN bracis-19048 367 46 hesburgh hesburgh PROPN bracis-19048 367 47 library library NOUN bracis-19048 367 48 ( ( PUNCT bracis-19048 367 49 3000184373 3000184373 NUM bracis-19048 367 50 ) ) PUNCT bracis-19048 368 1 © © ADP bracis-19048 368 2 2025 2025 NUM bracis-19048 368 3 springer springer NOUN bracis-19048 368 4 nature nature NOUN