id sid tid token lemma pos bracis-28409 1 1 flexcon flexcon NOUN bracis-28409 1 2 - - PUNCT bracis-28409 1 3 ce ce NOUN bracis-28409 1 4 : : PUNCT bracis-28409 1 5 a a DET bracis-28409 1 6 semi semi ADJ bracis-28409 1 7 - - ADJ bracis-28409 1 8 supervised supervised ADJ bracis-28409 1 9 method method NOUN bracis-28409 1 10 with with ADP bracis-28409 1 11     SPACE bracis-28409 1 12 an an DET bracis-28409 1 13     SPACE bracis-28409 1 14 ensemble ensemble ADJ bracis-28409 1 15 - - PUNCT bracis-28409 1 16 based base VERB bracis-28409 1 17 adaptive adaptive ADJ bracis-28409 1 18 confidence confidence NOUN bracis-28409 1 19 | | NOUN bracis-28409 1 20 springer springer NOUN bracis-28409 1 21 nature nature PROPN bracis-28409 1 22 link link PROPN bracis-28409 1 23 ( ( PUNCT bracis-28409 1 24 formerly formerly ADV bracis-28409 1 25 springerlink springerlink NOUN bracis-28409 1 26 ) ) PUNCT bracis-28409 1 27 skip skip VERB bracis-28409 1 28 to to ADP bracis-28409 1 29 main main ADJ bracis-28409 1 30 content content NOUN bracis-28409 1 31 advertisement advertisement NOUN bracis-28409 1 32 log log NOUN bracis-28409 1 33 in in ADP bracis-28409 1 34 menu menu NOUN bracis-28409 1 35 find find VERB bracis-28409 1 36 a a DET bracis-28409 1 37 journal journal NOUN bracis-28409 1 38 publish publish VERB bracis-28409 1 39 with with ADP bracis-28409 1 40 us we PRON bracis-28409 1 41 track track VERB bracis-28409 1 42 your your PRON bracis-28409 1 43 research research NOUN bracis-28409 1 44 search search NOUN bracis-28409 1 45 cart cart NOUN bracis-28409 1 46 home home NOUN bracis-28409 1 47 intelligent intelligent ADJ bracis-28409 1 48 systems system NOUN bracis-28409 1 49 conference conference NOUN bracis-28409 1 50 paper paper NOUN bracis-28409 1 51 flexcon flexcon NOUN bracis-28409 1 52 - - PUNCT bracis-28409 1 53 ce ce NOUN bracis-28409 1 54 : : PUNCT bracis-28409 1 55 a a DET bracis-28409 1 56 semi semi ADJ bracis-28409 1 57 - - ADJ bracis-28409 1 58 supervised supervised ADJ bracis-28409 1 59 method method NOUN bracis-28409 1 60 with with ADP bracis-28409 1 61     SPACE bracis-28409 1 62 an an DET bracis-28409 1 63     SPACE bracis-28409 1 64 ensemble ensemble ADJ bracis-28409 1 65 - - PUNCT bracis-28409 1 66 based base VERB bracis-28409 1 67 adaptive adaptive ADJ bracis-28409 1 68 confidence confidence NOUN bracis-28409 1 69 conference conference NOUN bracis-28409 1 70 paper paper NOUN bracis-28409 1 71 first first ADV bracis-28409 1 72 online online ADV bracis-28409 2 1 : : PUNCT bracis-28409 2 2 12 12 NUM bracis-28409 2 3 october october NOUN bracis-28409 2 4 2023 2023 NUM bracis-28409 3 1 pp pp PROPN bracis-28409 3 2 95–109 95–109 PROPN bracis-28409 3 3 cite cite VERB bracis-28409 3 4 this this DET bracis-28409 3 5 conference conference NOUN bracis-28409 3 6 paper paper NOUN bracis-28409 3 7 access access NOUN bracis-28409 3 8 provided provide VERB bracis-28409 3 9 by by ADP bracis-28409 3 10 university university PROPN bracis-28409 3 11 of of ADP bracis-28409 3 12 notre notre PROPN bracis-28409 3 13 dame dame PROPN bracis-28409 3 14 hesburgh hesburgh PROPN bracis-28409 3 15 library library PROPN bracis-28409 3 16 download download PROPN bracis-28409 3 17 book book NOUN bracis-28409 3 18 pdf pdf PROPN bracis-28409 3 19 download download NOUN bracis-28409 3 20 book book NOUN bracis-28409 3 21 epub epub PROPN bracis-28409 3 22 intelligent intelligent ADJ bracis-28409 3 23 systems system NOUN bracis-28409 3 24 ( ( PUNCT bracis-28409 3 25 bracis bracis NOUN bracis-28409 3 26 2023 2023 NUM bracis-28409 3 27 ) ) PUNCT bracis-28409 3 28 flexcon flexcon NOUN bracis-28409 3 29 - - PUNCT bracis-28409 3 30 ce ce PROPN bracis-28409 3 31 : : PUNCT bracis-28409 3 32 a a DET bracis-28409 3 33 semi semi ADJ bracis-28409 3 34 - - ADJ bracis-28409 3 35 supervised supervised ADJ bracis-28409 3 36 method method NOUN bracis-28409 3 37 with with ADP bracis-28409 3 38     SPACE bracis-28409 3 39 an an DET bracis-28409 3 40     SPACE bracis-28409 3 41 ensemble ensemble ADJ bracis-28409 3 42 - - PUNCT bracis-28409 3 43 based base VERB bracis-28409 3 44 adaptive adaptive ADJ bracis-28409 3 45 confidence confidence NOUN bracis-28409 3 46 download download NOUN bracis-28409 3 47 book book NOUN bracis-28409 3 48 pdf pdf PROPN bracis-28409 3 49 download download NOUN bracis-28409 3 50 book book PROPN bracis-28409 3 51 epub epub PROPN bracis-28409 3 52 arthur arthur PROPN bracis-28409 3 53 medeiros medeiros PROPN bracis-28409 3 54     SPACE bracis-28409 3 55 orcid orcid NOUN bracis-28409 3 56 : : PUNCT bracis-28409 3 57 orcid.org/0009-0003-3716-30229 orcid.org/0009-0003-3716-30229 NOUN bracis-28409 3 58 , , PUNCT bracis-28409 3 59 arthur arthur PROPN bracis-28409 3 60 c. c. PROPN bracis-28409 3 61 gorgônio gorgônio PROPN bracis-28409 3 62     SPACE bracis-28409 3 63 orcid orcid NOUN bracis-28409 3 64 : : PUNCT bracis-28409 3 65 orcid.org/0000-0002-1824-960010 orcid.org/0000-0002-1824-960010 NOUN bracis-28409 3 66 , , PUNCT bracis-28409 3 67 karliane karliane PROPN bracis-28409 3 68 medeiros medeiros PROPN bracis-28409 3 69 ovidio ovidio ADJ bracis-28409 3 70 vale vale ADJ bracis-28409 3 71     SPACE bracis-28409 3 72 orcid orcid NOUN bracis-28409 3 73 : : PUNCT bracis-28409 3 74 orcid.org/0000-0001-9845-81569 orcid.org/0000-0001-9845-81569 NOUN bracis-28409 3 75 , , PUNCT bracis-28409 3 76 flavius flavius PROPN bracis-28409 3 77 l. l. PROPN bracis-28409 3 78 gorgônio gorgônio PROPN bracis-28409 3 79     SPACE bracis-28409 3 80 orcid orcid NOUN bracis-28409 3 81 : : PUNCT bracis-28409 3 82 orcid.org/0000-0002-4926-39909 orcid.org/0000-0002-4926-39909 ADJ bracis-28409 3 83 & & CCONJ bracis-28409 3 84 … … PUNCT bracis-28409 3 85 anne anne PROPN bracis-28409 3 86 magály magály PROPN bracis-28409 3 87 de de PROPN bracis-28409 3 88 paula paula PROPN bracis-28409 3 89 canuto canuto PROPN bracis-28409 3 90     SPACE bracis-28409 3 91 orcid orcid NOUN bracis-28409 3 92 : : PUNCT bracis-28409 3 93 orcid.org/0000-0002-3684-38149 orcid.org/0000-0002-3684-38149 NOUN bracis-28409 3 94     SPACE bracis-28409 3 95 show show VERB bracis-28409 3 96 authors author NOUN bracis-28409 3 97 part part NOUN bracis-28409 3 98 of of ADP bracis-28409 3 99 the the DET bracis-28409 3 100 book book NOUN bracis-28409 3 101 series series NOUN bracis-28409 3 102 : : PUNCT bracis-28409 3 103 lecture lecture NOUN bracis-28409 3 104 notes note NOUN bracis-28409 3 105 in in ADP bracis-28409 3 106 computer computer NOUN bracis-28409 3 107 science science NOUN bracis-28409 3 108 ( ( PUNCT bracis-28409 3 109 ( ( PUNCT bracis-28409 3 110 lnai lnai ADJ bracis-28409 3 111 , , PUNCT bracis-28409 3 112 volume volume NOUN bracis-28409 3 113 14197 14197 NUM bracis-28409 3 114 ) ) PUNCT bracis-28409 3 115 ) ) PUNCT bracis-28409 3 116 included include VERB bracis-28409 3 117 in in ADP bracis-28409 3 118 the the DET bracis-28409 3 119 following follow VERB bracis-28409 3 120 conference conference NOUN bracis-28409 3 121 series series NOUN bracis-28409 3 122 : : PUNCT bracis-28409 3 123 brazilian brazilian ADJ bracis-28409 3 124 conference conference NOUN bracis-28409 3 125 on on ADP bracis-28409 3 126 intelligent intelligent ADJ bracis-28409 3 127 systems system NOUN bracis-28409 3 128 519 519 NUM bracis-28409 3 129 accesses access NOUN bracis-28409 3 130 2 2 NUM bracis-28409 3 131 citations citation NOUN bracis-28409 3 132 abstract abstract ADJ bracis-28409 3 133 semi semi ADJ bracis-28409 3 134 - - ADJ bracis-28409 3 135 supervised supervised ADJ bracis-28409 3 136 learning learning NOUN bracis-28409 3 137 is be AUX bracis-28409 3 138 characterized characterize VERB bracis-28409 3 139 by by ADP bracis-28409 3 140 a a DET bracis-28409 3 141 low low ADJ bracis-28409 3 142 number number NOUN bracis-28409 3 143 of of ADP bracis-28409 3 144 labeled label VERB bracis-28409 3 145 instances instance NOUN bracis-28409 3 146 and and CCONJ bracis-28409 3 147 a a DET bracis-28409 3 148 high high ADJ bracis-28409 3 149 number number NOUN bracis-28409 3 150 of of ADP bracis-28409 3 151 unlabeled unlabeled ADJ bracis-28409 3 152 instances instance NOUN bracis-28409 3 153 . . PUNCT bracis-28409 4 1 flexcon flexcon NOUN bracis-28409 4 2 - - PUNCT bracis-28409 4 3 c c NOUN bracis-28409 4 4 ( ( PUNCT bracis-28409 4 5 flexible flexible ADJ bracis-28409 4 6 confidence confidence NOUN bracis-28409 4 7 classifier classifier NOUN bracis-28409 4 8 ) ) PUNCT bracis-28409 4 9 is be AUX bracis-28409 4 10 a a DET bracis-28409 4 11 well well ADV bracis-28409 4 12 - - PUNCT bracis-28409 4 13 known know VERB bracis-28409 4 14 semi semi ADJ bracis-28409 4 15 - - ADJ bracis-28409 4 16 supervised supervised ADJ bracis-28409 4 17 method method NOUN bracis-28409 4 18 that that PRON bracis-28409 4 19 uses use VERB bracis-28409 4 20 the the DET bracis-28409 4 21 self self NOUN bracis-28409 4 22 - - PUNCT bracis-28409 4 23 training training NOUN bracis-28409 4 24 learning learn VERB bracis-28409 4 25 algorithm algorithm NOUN bracis-28409 4 26 as as ADP bracis-28409 4 27 basis basis NOUN bracis-28409 4 28 to to PART bracis-28409 4 29 generate generate VERB bracis-28409 4 30 prediction prediction NOUN bracis-28409 4 31 models model NOUN bracis-28409 4 32 . . PUNCT bracis-28409 5 1 the the DET bracis-28409 5 2 main main ADJ bracis-28409 5 3 difference difference NOUN bracis-28409 5 4 between between ADP bracis-28409 5 5 self self NOUN bracis-28409 5 6 - - PUNCT bracis-28409 5 7 training training NOUN bracis-28409 5 8 and and CCONJ bracis-28409 5 9 flexcon flexcon NOUN bracis-28409 5 10 - - PUNCT bracis-28409 5 11 c c NOUN bracis-28409 5 12 is be AUX bracis-28409 5 13 that that SCONJ bracis-28409 5 14 the the DET bracis-28409 5 15 former former ADJ bracis-28409 5 16 uses use VERB bracis-28409 5 17 a a DET bracis-28409 5 18 fixed fix VERB bracis-28409 5 19 threshold threshold NOUN bracis-28409 5 20 to to PART bracis-28409 5 21 select select VERB bracis-28409 5 22 the the DET bracis-28409 5 23 unlabeled unlabeled ADJ bracis-28409 5 24 instances instance NOUN bracis-28409 5 25 , , PUNCT bracis-28409 5 26 while while SCONJ bracis-28409 5 27 the the DET bracis-28409 5 28 latter latter NOUN bracis-28409 5 29 has have VERB bracis-28409 5 30 a a DET bracis-28409 5 31 dynamically dynamically ADV bracis-28409 5 32 adjusted adjust VERB bracis-28409 5 33 confidence confidence NOUN bracis-28409 5 34 . . PUNCT bracis-28409 6 1 flexcon flexcon NOUN bracis-28409 6 2 - - PUNCT bracis-28409 6 3 c c NOUN bracis-28409 6 4 applies apply VERB bracis-28409 6 5 a a DET bracis-28409 6 6 confidence confidence NOUN bracis-28409 6 7 adjustment adjustment NOUN bracis-28409 6 8 equation equation NOUN bracis-28409 6 9 based base VERB bracis-28409 6 10 on on ADP bracis-28409 6 11 the the DET bracis-28409 6 12 classifier classifier NOUN bracis-28409 6 13 performance performance NOUN bracis-28409 6 14 . . PUNCT bracis-28409 7 1 in in ADP bracis-28409 7 2 this this DET bracis-28409 7 3 sense sense NOUN bracis-28409 7 4 , , PUNCT bracis-28409 7 5 the the DET bracis-28409 7 6 classifier classifier NOUN bracis-28409 7 7 performance performance NOUN bracis-28409 7 8 is be AUX bracis-28409 7 9 used use VERB bracis-28409 7 10 to to PART bracis-28409 7 11 select select VERB bracis-28409 7 12 and and CCONJ bracis-28409 7 13 to to PART bracis-28409 7 14 label label VERB bracis-28409 7 15 unlabeled unlabele VERB bracis-28409 7 16 instances instance NOUN bracis-28409 7 17 . . PUNCT bracis-28409 8 1 in in ADP bracis-28409 8 2 machine machine NOUN bracis-28409 8 3 learning learning NOUN bracis-28409 8 4 , , PUNCT bracis-28409 8 5 it it PRON bracis-28409 8 6 is be AUX bracis-28409 8 7 well well ADV bracis-28409 8 8 - - PUNCT bracis-28409 8 9 known know VERB bracis-28409 8 10 that that SCONJ bracis-28409 8 11 the the DET bracis-28409 8 12 classifier classifier NOUN bracis-28409 8 13 performance performance NOUN bracis-28409 8 14 can can AUX bracis-28409 8 15 be be AUX bracis-28409 8 16 further far ADV bracis-28409 8 17 improved improve VERB bracis-28409 8 18 through through ADP bracis-28409 8 19 the the DET bracis-28409 8 20 use use NOUN bracis-28409 8 21 of of ADP bracis-28409 8 22 classifier classifier ADJ bracis-28409 8 23 ensembles ensemble NOUN bracis-28409 8 24 . . PUNCT bracis-28409 9 1 therefore therefore ADV bracis-28409 9 2 , , PUNCT bracis-28409 9 3 this this DET bracis-28409 9 4 study study NOUN bracis-28409 9 5 proposes propose VERB bracis-28409 9 6 the the DET bracis-28409 9 7 use use NOUN bracis-28409 9 8 classifier classifier NOUN bracis-28409 9 9 ensembles ensemble NOUN bracis-28409 9 10 in in ADP bracis-28409 9 11 the the DET bracis-28409 9 12 flexcon flexcon NOUN bracis-28409 9 13 - - PUNCT bracis-28409 9 14 c c NOUN bracis-28409 9 15 confidence confidence NOUN bracis-28409 9 16 adjustment adjustment NOUN bracis-28409 9 17 equation equation NOUN bracis-28409 9 18 , , PUNCT bracis-28409 9 19 aiming aim VERB bracis-28409 9 20 to to PART bracis-28409 9 21 provide provide VERB bracis-28409 9 22 a a DET bracis-28409 9 23 more more ADV bracis-28409 9 24 efficient efficient ADJ bracis-28409 9 25 measure measure NOUN bracis-28409 9 26 to to PART bracis-28409 9 27 select select VERB bracis-28409 9 28 and and CCONJ bracis-28409 9 29 to to PART bracis-28409 9 30 label label VERB bracis-28409 9 31 unlabeled unlabele VERB bracis-28409 9 32 instances instance NOUN bracis-28409 9 33 . . PUNCT bracis-28409 10 1 in in ADP bracis-28409 10 2 order order NOUN bracis-28409 10 3 to to PART bracis-28409 10 4 assess assess VERB bracis-28409 10 5 the the DET bracis-28409 10 6 viability viability NOUN bracis-28409 10 7 of of ADP bracis-28409 10 8 the the DET bracis-28409 10 9 proposed propose VERB bracis-28409 10 10 method method NOUN bracis-28409 10 11 ( ( PUNCT bracis-28409 10 12 flexcon flexcon NOUN bracis-28409 10 13 - - PUNCT bracis-28409 10 14 ce ce NOUN bracis-28409 10 15 ) ) PUNCT bracis-28409 10 16 , , PUNCT bracis-28409 10 17 an an DET bracis-28409 10 18 empirical empirical ADJ bracis-28409 10 19 analysis analysis NOUN bracis-28409 10 20 will will AUX bracis-28409 10 21 be be AUX bracis-28409 10 22 conducted conduct VERB bracis-28409 10 23 , , PUNCT bracis-28409 10 24 using use VERB bracis-28409 10 25 20 20 NUM bracis-28409 10 26 datasets dataset NOUN bracis-28409 10 27 , , PUNCT bracis-28409 10 28 three three NUM bracis-28409 10 29 different different ADJ bracis-28409 10 30 classification classification NOUN bracis-28409 10 31 algorithms algorithm NOUN bracis-28409 10 32 and and CCONJ bracis-28409 10 33 five five NUM bracis-28409 10 34 different different ADJ bracis-28409 10 35 configurations configuration NOUN bracis-28409 10 36 of of ADP bracis-28409 10 37 initially initially ADV bracis-28409 10 38 unlabeled unlabele VERB bracis-28409 10 39 data datum NOUN bracis-28409 10 40 . . PUNCT bracis-28409 11 1 the the DET bracis-28409 11 2 results result NOUN bracis-28409 11 3 indicate indicate VERB bracis-28409 11 4 that that SCONJ bracis-28409 11 5 the the DET bracis-28409 11 6 proposed propose VERB bracis-28409 11 7 method method NOUN bracis-28409 11 8 outperformed outperform VERB bracis-28409 11 9 the the DET bracis-28409 11 10 traditional traditional ADJ bracis-28409 11 11 method method NOUN bracis-28409 11 12 , , PUNCT bracis-28409 11 13 therewith therewith NOUN bracis-28409 11 14 proving prove VERB bracis-28409 11 15 itself itself PRON bracis-28409 11 16 promising promise VERB bracis-28409 11 17 for for ADP bracis-28409 11 18 the the DET bracis-28409 11 19 task task NOUN bracis-28409 11 20 of of ADP bracis-28409 11 21 automatic automatic ADJ bracis-28409 11 22 data datum NOUN bracis-28409 11 23 selection selection NOUN bracis-28409 11 24 and and CCONJ bracis-28409 11 25 labeling labeling NOUN bracis-28409 11 26 in in ADP bracis-28409 11 27 the the DET bracis-28409 11 28 semi semi ADJ bracis-28409 11 29 - - ADJ bracis-28409 11 30 supervised supervised ADJ bracis-28409 11 31 context context NOUN bracis-28409 11 32 . . PUNCT bracis-28409 12 1 access access NOUN bracis-28409 12 2 provided provide VERB bracis-28409 12 3 by by ADP bracis-28409 12 4 university university PROPN bracis-28409 12 5 of of ADP bracis-28409 12 6 notre notre PROPN bracis-28409 12 7 dame dame PROPN bracis-28409 12 8 hesburgh hesburgh PROPN bracis-28409 12 9 library library PROPN bracis-28409 12 10 . . PUNCT bracis-28409 13 1 download download PROPN bracis-28409 13 2 conference conference NOUN bracis-28409 13 3 paper paper NOUN bracis-28409 13 4 pdf pdf NOUN bracis-28409 13 5 similar similar ADJ bracis-28409 13 6 content content NOUN bracis-28409 13 7 being be AUX bracis-28409 13 8 viewed view VERB bracis-28409 13 9 by by ADP bracis-28409 13 10 others other NOUN bracis-28409 13 11 ensemble ensemble ADJ bracis-28409 13 12 enhanced enhance VERB bracis-28409 13 13 active active ADJ bracis-28409 13 14 learning learning NOUN bracis-28409 13 15 mixture mixture NOUN bracis-28409 13 16 discriminant discriminant NOUN bracis-28409 13 17 analysis analysis NOUN bracis-28409 13 18 model model NOUN bracis-28409 13 19 and and CCONJ bracis-28409 13 20 its its PRON bracis-28409 13 21 application application NOUN bracis-28409 13 22 for for ADP bracis-28409 13 23 semi semi ADJ bracis-28409 13 24 - - ADJ bracis-28409 13 25 supervised supervised ADJ bracis-28409 13 26 fault fault NOUN bracis-28409 13 27 classification classification NOUN bracis-28409 13 28 article article NOUN bracis-28409 13 29 26 26 NUM bracis-28409 13 30 july july PROPN bracis-28409 13 31 2022 2022 NUM bracis-28409 13 32 semi semi ADV bracis-28409 13 33 - - ADJ bracis-28409 13 34 supervised supervised ADJ bracis-28409 13 35 learning learning NOUN bracis-28409 13 36 chapter chapter NOUN bracis-28409 14 1 © © PROPN bracis-28409 14 2 2025 2025 NUM bracis-28409 14 3 learning learn VERB bracis-28409 14 4 sample sample NOUN bracis-28409 14 5 - - PUNCT bracis-28409 14 6 aware aware ADJ bracis-28409 14 7 threshold threshold NOUN bracis-28409 14 8 for for ADP bracis-28409 14 9 semi semi ADJ bracis-28409 14 10 - - ADJ bracis-28409 14 11 supervised supervised ADJ bracis-28409 14 12 learning learning NOUN bracis-28409 14 13 article article NOUN bracis-28409 14 14 18 18 NUM bracis-28409 14 15 january january PROPN bracis-28409 14 16 2024 2024 NUM bracis-28409 14 17 explore explore VERB bracis-28409 14 18 related relate VERB bracis-28409 14 19 subjects subject NOUN bracis-28409 14 20 discover discover VERB bracis-28409 14 21 the the DET bracis-28409 14 22 latest late ADJ bracis-28409 14 23 articles article NOUN bracis-28409 14 24 , , PUNCT bracis-28409 14 25 books book NOUN bracis-28409 14 26 and and CCONJ bracis-28409 14 27 news news NOUN bracis-28409 14 28 in in ADP bracis-28409 14 29 related related ADJ bracis-28409 14 30 subjects subject NOUN bracis-28409 14 31 , , PUNCT bracis-28409 14 32 suggested suggest VERB bracis-28409 14 33 using use VERB bracis-28409 14 34 machine machine NOUN bracis-28409 14 35 learning learning NOUN bracis-28409 14 36 . . PUNCT bracis-28409 15 1 categorization categorization NOUN bracis-28409 15 2 classical classical ADJ bracis-28409 15 3 conditioning conditioning NOUN bracis-28409 15 4 learning learn VERB bracis-28409 15 5 algorithms algorithm NOUN bracis-28409 15 6 machine machine NOUN bracis-28409 15 7 learning learn VERB bracis-28409 15 8 sensory sensory ADJ bracis-28409 15 9 evaluation evaluation NOUN bracis-28409 15 10 statistical statistical ADJ bracis-28409 15 11 learning learning NOUN bracis-28409 15 12 1 1 NUM bracis-28409 15 13 introduction introduction NOUN bracis-28409 15 14 known know VERB bracis-28409 15 15 as as ADP bracis-28409 15 16 a a DET bracis-28409 15 17 field field NOUN bracis-28409 15 18 of of ADP bracis-28409 15 19 study study NOUN bracis-28409 15 20 that that PRON bracis-28409 15 21 has have AUX bracis-28409 15 22 gained gain VERB bracis-28409 15 23 significant significant ADJ bracis-28409 15 24 importance importance NOUN bracis-28409 15 25 in in ADP bracis-28409 15 26 recent recent ADJ bracis-28409 15 27 years year NOUN bracis-28409 15 28 , , PUNCT bracis-28409 15 29 machine machine NOUN bracis-28409 15 30 learning learning NOUN bracis-28409 15 31 ( ( PUNCT bracis-28409 15 32 ml ml NOUN bracis-28409 15 33 ) ) PUNCT bracis-28409 15 34 has have AUX bracis-28409 15 35 emerged emerge VERB bracis-28409 15 36 as as ADP bracis-28409 15 37 a a DET bracis-28409 15 38 way way NOUN bracis-28409 15 39 of of ADP bracis-28409 15 40 deducing deduce VERB bracis-28409 15 41 a a DET bracis-28409 15 42 hypothesis hypothesis NOUN bracis-28409 15 43 based base VERB bracis-28409 15 44 on on ADP bracis-28409 15 45 past past NOUN bracis-28409 15 46 or or CCONJ bracis-28409 15 47 lived live VERB bracis-28409 15 48 experience experience NOUN bracis-28409 15 49 [ [ X bracis-28409 15 50 1 1 NUM bracis-28409 15 51 ] ] PUNCT bracis-28409 15 52 . . PUNCT bracis-28409 16 1 in in ADP bracis-28409 16 2 this this DET bracis-28409 16 3 context context NOUN bracis-28409 16 4 , , PUNCT bracis-28409 16 5 machines machine NOUN bracis-28409 16 6 are be AUX bracis-28409 16 7 programmed program VERB bracis-28409 16 8 to to PART bracis-28409 16 9 learn learn VERB bracis-28409 16 10 from from ADP bracis-28409 16 11 previously previously ADV bracis-28409 16 12 stored store VERB bracis-28409 16 13 data datum NOUN bracis-28409 16 14 derived derive VERB bracis-28409 16 15 from from ADP bracis-28409 16 16 previous previous ADJ bracis-28409 16 17 experiences experience NOUN bracis-28409 16 18 in in ADP bracis-28409 16 19 order order NOUN bracis-28409 16 20 to to PART bracis-28409 16 21 generate generate VERB bracis-28409 16 22 knowledge knowledge NOUN bracis-28409 16 23 . . PUNCT bracis-28409 17 1 in in ADP bracis-28409 17 2 ml ml ADP bracis-28409 17 3 , , PUNCT bracis-28409 17 4 the the DET bracis-28409 17 5 learning learn VERB bracis-28409 17 6 algorithms algorithm NOUN bracis-28409 17 7 can can AUX bracis-28409 17 8 be be AUX bracis-28409 17 9 broadly broadly ADV bracis-28409 17 10 divided divide VERB bracis-28409 17 11 into into ADP bracis-28409 17 12 supervised supervised ADJ bracis-28409 17 13 , , PUNCT bracis-28409 17 14 semi semi ADJ bracis-28409 17 15 - - ADJ bracis-28409 17 16 supervised supervised ADJ bracis-28409 17 17 and and CCONJ bracis-28409 17 18 unsupervised unsupervised ADJ bracis-28409 17 19 algorithms algorithm NOUN bracis-28409 17 20 . . PUNCT bracis-28409 18 1 in in ADP bracis-28409 18 2 this this DET bracis-28409 18 3 paper paper NOUN bracis-28409 18 4 , , PUNCT bracis-28409 18 5 semi semi ADJ bracis-28409 18 6 - - ADJ bracis-28409 18 7 supervised supervised ADJ bracis-28409 18 8 learning learning NOUN bracis-28409 18 9 ( ( PUNCT bracis-28409 18 10 ssl ssl ADJ bracis-28409 18 11 ) ) PUNCT bracis-28409 18 12 algorithms algorithm NOUN bracis-28409 18 13 are be AUX bracis-28409 18 14 assessed assess VERB bracis-28409 18 15 and and CCONJ bracis-28409 18 16 these these DET bracis-28409 18 17 algorithms algorithm NOUN bracis-28409 18 18 use use VERB bracis-28409 18 19 labeled label VERB bracis-28409 18 20 instances instance NOUN bracis-28409 18 21 to to PART bracis-28409 18 22 build build VERB bracis-28409 18 23 their their PRON bracis-28409 18 24 initial initial ADJ bracis-28409 18 25 hypothesis hypothesis NOUN bracis-28409 18 26 and and CCONJ bracis-28409 18 27 combine combine VERB bracis-28409 18 28 the the DET bracis-28409 18 29 information information NOUN bracis-28409 18 30 obtained obtain VERB bracis-28409 18 31 by by ADP bracis-28409 18 32 these these DET bracis-28409 18 33 examples example NOUN bracis-28409 18 34 to to PART bracis-28409 18 35 label label VERB bracis-28409 18 36 the the DET bracis-28409 18 37 unlabeled unlabele VERB bracis-28409 18 38 instances instance NOUN bracis-28409 18 39 . . PUNCT bracis-28409 19 1 in in ADP bracis-28409 19 2 other other ADJ bracis-28409 19 3 words word NOUN bracis-28409 19 4 , , PUNCT bracis-28409 19 5 it it PRON bracis-28409 19 6 is be AUX bracis-28409 19 7 possible possible ADJ bracis-28409 19 8 to to PART bracis-28409 19 9 use use VERB bracis-28409 19 10 partially partially ADV bracis-28409 19 11 supervised supervise VERB bracis-28409 19 12 information information NOUN bracis-28409 19 13 to to PART bracis-28409 19 14 guide guide VERB bracis-28409 19 15 the the DET bracis-28409 19 16 learning learning NOUN bracis-28409 19 17 process process NOUN bracis-28409 19 18 and and CCONJ bracis-28409 19 19 increase increase VERB bracis-28409 19 20 the the DET bracis-28409 19 21 evidence evidence NOUN bracis-28409 19 22 of of ADP bracis-28409 19 23 the the DET bracis-28409 19 24 target target NOUN bracis-28409 19 25 labels label NOUN bracis-28409 19 26 [ [ X bracis-28409 19 27 2 2 NUM bracis-28409 19 28 ] ] PUNCT bracis-28409 19 29 . . PUNCT bracis-28409 20 1 nevertheless nevertheless ADV bracis-28409 20 2 , , PUNCT bracis-28409 20 3 we we PRON bracis-28409 20 4 have have AUX bracis-28409 20 5 observed observe VERB bracis-28409 20 6 that that SCONJ bracis-28409 20 7 the the DET bracis-28409 20 8 automatic automatic ADJ bracis-28409 20 9 assignment assignment NOUN bracis-28409 20 10 of of ADP bracis-28409 20 11 labels label NOUN bracis-28409 20 12 in in ADP bracis-28409 20 13 an an DET bracis-28409 20 14 ssl ssl ADJ bracis-28409 20 15 algorithm algorithm NOUN bracis-28409 20 16 is be AUX bracis-28409 20 17 a a DET bracis-28409 20 18 difficult difficult ADJ bracis-28409 20 19 task task NOUN bracis-28409 20 20 , , PUNCT bracis-28409 20 21 mainly mainly ADV bracis-28409 20 22 related relate VERB bracis-28409 20 23 to to ADP bracis-28409 20 24 the the DET bracis-28409 20 25 correct correct ADJ bracis-28409 20 26 selection selection NOUN bracis-28409 20 27 of of ADP bracis-28409 20 28 the the DET bracis-28409 20 29 unlabeled unlabele VERB bracis-28409 20 30 instances instance NOUN bracis-28409 20 31 to to PART bracis-28409 20 32 be be AUX bracis-28409 20 33 labeled label VERB bracis-28409 20 34 . . PUNCT bracis-28409 21 1 an an DET bracis-28409 21 2 inefficient inefficient ADJ bracis-28409 21 3 selection selection NOUN bracis-28409 21 4 of of ADP bracis-28409 21 5 the the DET bracis-28409 21 6 unlabeled unlabele VERB bracis-28409 21 7 instances instance NOUN bracis-28409 21 8 and and CCONJ bracis-28409 21 9 the the DET bracis-28409 21 10 consequent consequent ADJ bracis-28409 21 11 labeling labeling NOUN bracis-28409 21 12 of of ADP bracis-28409 21 13 these these DET bracis-28409 21 14 instances instance NOUN bracis-28409 21 15 can can AUX bracis-28409 21 16 degrade degrade VERB bracis-28409 21 17 enormously enormously ADV bracis-28409 21 18 the the DET bracis-28409 21 19 performance performance NOUN bracis-28409 21 20 of of ADP bracis-28409 21 21 the the DET bracis-28409 21 22 ssl ssl ADJ bracis-28409 21 23 algorithms algorithm NOUN bracis-28409 21 24 . . PUNCT bracis-28409 22 1 a a DET bracis-28409 22 2 solution solution NOUN bracis-28409 22 3 for for ADP bracis-28409 22 4 this this DET bracis-28409 22 5 problem problem NOUN bracis-28409 22 6 is be AUX bracis-28409 22 7 proposed propose VERB bracis-28409 22 8 in in ADP bracis-28409 22 9 [ [ X bracis-28409 22 10 3 3 X bracis-28409 22 11 ] ] PUNCT bracis-28409 22 12 using use VERB bracis-28409 22 13 a a DET bracis-28409 22 14 semi semi ADJ bracis-28409 22 15 - - ADJ bracis-28409 22 16 supervised supervised ADJ bracis-28409 22 17 learning learning NOUN bracis-28409 22 18 model model NOUN bracis-28409 22 19 called call VERB bracis-28409 22 20 flexible flexible ADJ bracis-28409 22 21 confidence confidence NOUN bracis-28409 22 22 classifier classifier NOUN bracis-28409 22 23 ( ( PUNCT bracis-28409 22 24 flexcon flexcon NOUN bracis-28409 22 25 - - PUNCT bracis-28409 22 26 c c NOUN bracis-28409 22 27 ) ) PUNCT bracis-28409 22 28 , , PUNCT bracis-28409 22 29 which which PRON bracis-28409 22 30 automatically automatically ADV bracis-28409 22 31 adjusts adjust VERB bracis-28409 22 32 the the DET bracis-28409 22 33 confidence confidence NOUN bracis-28409 22 34 parameter parameter NOUN bracis-28409 22 35 to to PART bracis-28409 22 36 select select VERB bracis-28409 22 37 and and CCONJ bracis-28409 22 38 label label VERB bracis-28409 22 39 the the DET bracis-28409 22 40 unlabeled unlabeled ADJ bracis-28409 22 41 instance instance NOUN bracis-28409 22 42 . . PUNCT bracis-28409 23 1 this this DET bracis-28409 23 2 confidence confidence NOUN bracis-28409 23 3 is be AUX bracis-28409 23 4 based base VERB bracis-28409 23 5 solely solely ADV bracis-28409 23 6 on on ADP bracis-28409 23 7 the the DET bracis-28409 23 8 accuracy accuracy NOUN bracis-28409 23 9 of of ADP bracis-28409 23 10 a a DET bracis-28409 23 11 classifier classifier NOUN bracis-28409 23 12 built build VERB bracis-28409 23 13 in in ADP bracis-28409 23 14 the the DET bracis-28409 23 15 previous previous ADJ bracis-28409 23 16 iterations iteration NOUN bracis-28409 23 17 . . PUNCT bracis-28409 24 1 in in ADP bracis-28409 24 2 the the DET bracis-28409 24 3 ml ml NOUN bracis-28409 24 4 context context NOUN bracis-28409 24 5 , , PUNCT bracis-28409 24 6 it it PRON bracis-28409 24 7 is be AUX bracis-28409 24 8 well well ADV bracis-28409 24 9 - - PUNCT bracis-28409 24 10 known know VERB bracis-28409 24 11 that that SCONJ bracis-28409 24 12 the the DET bracis-28409 24 13 accuracy accuracy NOUN bracis-28409 24 14 of of ADP bracis-28409 24 15 a a DET bracis-28409 24 16 single single ADJ bracis-28409 24 17 classifier classifier NOUN bracis-28409 24 18 can can AUX bracis-28409 24 19 not not PART bracis-28409 24 20 be be AUX bracis-28409 24 21 efficient efficient ADJ bracis-28409 24 22 for for ADP bracis-28409 24 23 complex complex ADJ bracis-28409 24 24 tasks task NOUN bracis-28409 24 25 . . PUNCT bracis-28409 25 1 for for ADP bracis-28409 25 2 this this DET bracis-28409 25 3 reason reason NOUN bracis-28409 25 4 , , PUNCT bracis-28409 25 5 combining combine VERB bracis-28409 25 6 different different ADJ bracis-28409 25 7 classifiers classifier NOUN bracis-28409 25 8 to to PART bracis-28409 25 9 improve improve VERB bracis-28409 25 10 the the DET bracis-28409 25 11 classifier classifier NOUN bracis-28409 25 12 performance performance NOUN bracis-28409 25 13 has have AUX bracis-28409 25 14 been be AUX bracis-28409 25 15 emerged emerge VERB bracis-28409 25 16 as as ADP bracis-28409 25 17 an an DET bracis-28409 25 18 efficient efficient ADJ bracis-28409 25 19 alternative alternative NOUN bracis-28409 25 20 , , PUNCT bracis-28409 25 21 as as SCONJ bracis-28409 25 22 observed observe VERB bracis-28409 25 23 in in ADP bracis-28409 25 24 several several ADJ bracis-28409 25 25 studies study NOUN bracis-28409 25 26 in in ADP bracis-28409 25 27 the the DET bracis-28409 25 28 field field NOUN bracis-28409 26 1 [ [ X bracis-28409 26 2 4,5,6,7,8,9 4,5,6,7,8,9 NUM bracis-28409 26 3 ] ] PUNCT bracis-28409 26 4 . . PUNCT bracis-28409 27 1 in in ADP bracis-28409 27 2 these these DET bracis-28409 27 3 studies study NOUN bracis-28409 27 4 , , PUNCT bracis-28409 27 5 it it PRON bracis-28409 27 6 has have AUX bracis-28409 27 7 been be AUX bracis-28409 27 8 noticed notice VERB bracis-28409 27 9 that that SCONJ bracis-28409 27 10 combining combine VERB bracis-28409 27 11 information information NOUN bracis-28409 27 12 from from ADP bracis-28409 27 13 various various ADJ bracis-28409 27 14 sources source NOUN bracis-28409 27 15 has have AUX bracis-28409 27 16 resulted result VERB bracis-28409 27 17 in in ADP bracis-28409 27 18 systems system NOUN bracis-28409 27 19 with with ADP bracis-28409 27 20 more more ADJ bracis-28409 27 21 reliability reliability NOUN bracis-28409 27 22 and and CCONJ bracis-28409 27 23 performance performance NOUN bracis-28409 27 24 than than ADP bracis-28409 27 25 individual individual ADJ bracis-28409 27 26 classifiers classifier NOUN bracis-28409 27 27 . . PUNCT bracis-28409 28 1 systems system NOUN bracis-28409 28 2 combining combine VERB bracis-28409 28 3 classifiers classifier NOUN bracis-28409 28 4 are be AUX bracis-28409 28 5 known know VERB bracis-28409 28 6 as as ADP bracis-28409 28 7 classifier classifier NOUN bracis-28409 28 8 ensembles ensemble NOUN bracis-28409 28 9 and and CCONJ bracis-28409 28 10 they they PRON bracis-28409 28 11 are be AUX bracis-28409 28 12 composed compose VERB bracis-28409 28 13 of of ADP bracis-28409 28 14 a a DET bracis-28409 28 15 set set NOUN bracis-28409 28 16 of of ADP bracis-28409 28 17 base base NOUN bracis-28409 28 18 classifiers classifier NOUN bracis-28409 28 19 organized organize VERB bracis-28409 28 20 in in ADP bracis-28409 28 21 a a DET bracis-28409 28 22 parallel parallel ADJ bracis-28409 28 23 way way NOUN bracis-28409 28 24 . . PUNCT bracis-28409 29 1 these these DET bracis-28409 29 2 classifiers classifier NOUN bracis-28409 29 3 receive receive VERB bracis-28409 29 4 the the DET bracis-28409 29 5 input input NOUN bracis-28409 29 6 data datum NOUN bracis-28409 29 7 and and CCONJ bracis-28409 29 8 they they PRON bracis-28409 29 9 produce produce VERB bracis-28409 29 10 several several ADJ bracis-28409 29 11 outputs output NOUN bracis-28409 29 12 that that PRON bracis-28409 29 13 are be AUX bracis-28409 29 14 used use VERB bracis-28409 29 15 , , PUNCT bracis-28409 29 16 in in ADP bracis-28409 29 17 turn turn NOUN bracis-28409 29 18 , , PUNCT bracis-28409 29 19 as as ADP bracis-28409 29 20 inputs input NOUN bracis-28409 29 21 by by ADP bracis-28409 29 22 a a DET bracis-28409 29 23 combination combination NOUN bracis-28409 29 24 method method NOUN bracis-28409 29 25 , , PUNCT bracis-28409 29 26 which which PRON bracis-28409 29 27 combines combine VERB bracis-28409 29 28 these these DET bracis-28409 29 29 outputs output NOUN bracis-28409 29 30 and and CCONJ bracis-28409 29 31 generates generate VERB bracis-28409 29 32 the the DET bracis-28409 29 33 final final ADJ bracis-28409 29 34 ensemble ensemble ADJ bracis-28409 29 35 output output NOUN bracis-28409 29 36 [ [ X bracis-28409 29 37 7 7 NUM bracis-28409 29 38 ] ] PUNCT bracis-28409 29 39 . . PUNCT bracis-28409 30 1 in in ADP bracis-28409 30 2 this this DET bracis-28409 30 3 paper paper NOUN bracis-28409 30 4 , , PUNCT bracis-28409 30 5 we we PRON bracis-28409 30 6 seek seek VERB bracis-28409 30 7 to to PART bracis-28409 30 8 improve improve VERB bracis-28409 30 9 the the DET bracis-28409 30 10 flexcon flexcon NOUN bracis-28409 30 11 - - PUNCT bracis-28409 30 12 c c NOUN bracis-28409 31 1 [ [ X bracis-28409 31 2 3 3 NUM bracis-28409 31 3 ] ] PUNCT bracis-28409 31 4 performance performance NOUN bracis-28409 31 5 by by ADP bracis-28409 31 6 introducing introduce VERB bracis-28409 31 7 the the DET bracis-28409 31 8 classifier classifier NOUN bracis-28409 31 9 ensembles ensemble NOUN bracis-28409 31 10 strategy strategy NOUN bracis-28409 31 11 to to PART bracis-28409 31 12 select select VERB bracis-28409 31 13 and and CCONJ bracis-28409 31 14 to to PART bracis-28409 31 15 label label VERB bracis-28409 31 16 unlabeled unlabele VERB bracis-28409 31 17 instances instance NOUN bracis-28409 31 18 and and CCONJ bracis-28409 31 19 to to PART bracis-28409 31 20 include include VERB bracis-28409 31 21 them they PRON bracis-28409 31 22 in in ADP bracis-28409 31 23 the the DET bracis-28409 31 24 labeled label VERB bracis-28409 31 25 data datum NOUN bracis-28409 31 26 set set VERB bracis-28409 31 27 . . PUNCT bracis-28409 32 1 therefore therefore ADV bracis-28409 32 2 , , PUNCT bracis-28409 32 3 this this DET bracis-28409 32 4 paper paper NOUN bracis-28409 32 5 proposes propose VERB bracis-28409 32 6 the the DET bracis-28409 32 7 flexcon flexcon NOUN bracis-28409 32 8 - - PUNCT bracis-28409 32 9 ce ce PROPN bracis-28409 32 10 ( ( PUNCT bracis-28409 32 11 flexible flexible ADJ bracis-28409 32 12 confidence confidence NOUN bracis-28409 32 13 with with ADP bracis-28409 32 14 classifier classifier ADJ bracis-28409 32 15 ensembles ensemble NOUN bracis-28409 32 16 ) ) PUNCT bracis-28409 32 17 model model NOUN bracis-28409 32 18 , , PUNCT bracis-28409 32 19 which which PRON bracis-28409 32 20 uses use VERB bracis-28409 32 21 an an DET bracis-28409 32 22 ensemble ensemble ADJ bracis-28409 32 23 - - PUNCT bracis-28409 32 24 based base VERB bracis-28409 32 25 confidence confidence NOUN bracis-28409 32 26 adaptive adaptive ADJ bracis-28409 32 27 confidence confidence NOUN bracis-28409 32 28 equation equation NOUN bracis-28409 32 29 . . PUNCT bracis-28409 33 1 in in ADP bracis-28409 33 2 fact fact NOUN bracis-28409 33 3 , , PUNCT bracis-28409 33 4 two two NUM bracis-28409 33 5 different different ADJ bracis-28409 33 6 versions version NOUN bracis-28409 33 7 of of ADP bracis-28409 33 8 the the DET bracis-28409 33 9 flexcon flexcon NOUN bracis-28409 33 10 - - PUNCT bracis-28409 33 11 ce ce PROPN bracis-28409 33 12 method method NOUN bracis-28409 33 13 are be AUX bracis-28409 33 14 proposed propose VERB bracis-28409 33 15 in in ADP bracis-28409 33 16 this this DET bracis-28409 33 17 paper paper NOUN bracis-28409 33 18 . . PUNCT bracis-28409 34 1 the the DET bracis-28409 34 2 first first ADJ bracis-28409 34 3 one one NUM bracis-28409 34 4 , , PUNCT bracis-28409 34 5 called call VERB bracis-28409 34 6 flexcon flexcon NOUN bracis-28409 34 7 - - PUNCT bracis-28409 34 8 ce(hom ce(hom NOUN bracis-28409 34 9 ) ) PUNCT bracis-28409 34 10 , , PUNCT bracis-28409 34 11 uses use VERB bracis-28409 34 12 a a DET bracis-28409 34 13 homogeneous homogeneous ADJ bracis-28409 34 14 ensemble ensemble ADJ bracis-28409 34 15 ( ( PUNCT bracis-28409 34 16 base base NOUN bracis-28409 34 17 classifiers classifier NOUN bracis-28409 34 18 of of ADP bracis-28409 34 19 the the DET bracis-28409 34 20 same same ADJ bracis-28409 34 21 type type NOUN bracis-28409 34 22 ) ) PUNCT bracis-28409 34 23 while while SCONJ bracis-28409 34 24 the the DET bracis-28409 34 25 second second ADJ bracis-28409 34 26 one one NOUN bracis-28409 34 27 , , PUNCT bracis-28409 34 28 named name VERB bracis-28409 34 29 flexcon flexcon NOUN bracis-28409 34 30 - - PUNCT bracis-28409 34 31 ce(het ce(het NOUN bracis-28409 34 32 ) ) PUNCT bracis-28409 34 33 , , PUNCT bracis-28409 34 34 uses use VERB bracis-28409 34 35 an an DET bracis-28409 34 36 heterogeneous heterogeneous ADJ bracis-28409 34 37 ensemble ensemble NOUN bracis-28409 34 38 . . PUNCT bracis-28409 35 1 this this DET bracis-28409 35 2 study study NOUN bracis-28409 35 3 assesses assess VERB bracis-28409 35 4 the the DET bracis-28409 35 5 efficiency efficiency NOUN bracis-28409 35 6 of of ADP bracis-28409 35 7 the the DET bracis-28409 35 8 proposed propose VERB bracis-28409 35 9 methods method NOUN bracis-28409 35 10 when when SCONJ bracis-28409 35 11 building build VERB bracis-28409 35 12 representative representative ADJ bracis-28409 35 13 classifier classifier NOUN bracis-28409 35 14 ensembles ensemble NOUN bracis-28409 35 15 by by ADP bracis-28409 35 16 analyzing analyze VERB bracis-28409 35 17 the the DET bracis-28409 35 18 classification classification NOUN bracis-28409 35 19 effectiveness effectiveness NOUN bracis-28409 35 20 of of ADP bracis-28409 35 21 these these DET bracis-28409 35 22 methods method NOUN bracis-28409 35 23 in in ADP bracis-28409 35 24 20 20 NUM bracis-28409 35 25 well well ADV bracis-28409 35 26 - - PUNCT bracis-28409 35 27 known know VERB bracis-28409 35 28 datasets dataset NOUN bracis-28409 35 29 , , PUNCT bracis-28409 35 30 with with ADP bracis-28409 35 31 three three NUM bracis-28409 35 32 different different ADJ bracis-28409 35 33 classification classification NOUN bracis-28409 35 34 algorithms algorithm NOUN bracis-28409 35 35 and and CCONJ bracis-28409 35 36 five five NUM bracis-28409 35 37 different different ADJ bracis-28409 35 38 configurations configuration NOUN bracis-28409 35 39 of of ADP bracis-28409 35 40 initially initially ADV bracis-28409 35 41 unlabeled unlabele VERB bracis-28409 35 42 data datum NOUN bracis-28409 35 43 . . PUNCT bracis-28409 36 1 this this DET bracis-28409 36 2 article article NOUN bracis-28409 36 3 is be AUX bracis-28409 36 4 organized organize VERB bracis-28409 36 5 as as SCONJ bracis-28409 36 6 follows follow VERB bracis-28409 36 7 : : PUNCT bracis-28409 36 8 sect sect NOUN bracis-28409 36 9 . . PUNCT bracis-28409 36 10     SPACE bracis-28409 37 1 2 2 NUM bracis-28409 37 2 presents present VERB bracis-28409 37 3 the the DET bracis-28409 37 4 theoretical theoretical ADJ bracis-28409 37 5 aspects aspect NOUN bracis-28409 37 6 of of ADP bracis-28409 37 7 semi semi ADJ bracis-28409 37 8 - - ADJ bracis-28409 37 9 supervised supervised ADJ bracis-28409 37 10 learning learning NOUN bracis-28409 37 11 algorithms algorithm NOUN bracis-28409 37 12 as as ADV bracis-28409 37 13 well well ADV bracis-28409 37 14 as as ADP bracis-28409 37 15 the the DET bracis-28409 37 16 flexcon flexcon NOUN bracis-28409 37 17 - - PUNCT bracis-28409 37 18 c c NOUN bracis-28409 37 19 method method NOUN bracis-28409 37 20 and and CCONJ bracis-28409 37 21 the the DET bracis-28409 37 22 classifier classifier NOUN bracis-28409 37 23 ensembles ensemble NOUN bracis-28409 37 24 . . PUNCT bracis-28409 38 1 section section NOUN bracis-28409 38 2     SPACE bracis-28409 38 3 3 3 NUM bracis-28409 38 4 highlights highlight NOUN bracis-28409 38 5 some some DET bracis-28409 38 6 related related ADJ bracis-28409 38 7 work work NOUN bracis-28409 38 8 to to ADP bracis-28409 38 9 this this DET bracis-28409 38 10 research research NOUN bracis-28409 38 11 while while SCONJ bracis-28409 38 12 sect sect NOUN bracis-28409 38 13 . . PUNCT bracis-28409 38 14     SPACE bracis-28409 38 15 4 4 NUM bracis-28409 38 16 presents present NOUN bracis-28409 38 17 and and CCONJ bracis-28409 38 18 details detail NOUN bracis-28409 38 19 the the DET bracis-28409 38 20 ssl ssl PROPN bracis-28409 38 21 method method NOUN bracis-28409 38 22 proposed propose VERB bracis-28409 38 23 in in ADP bracis-28409 38 24 this this DET bracis-28409 38 25 article article NOUN bracis-28409 38 26 . . PUNCT bracis-28409 39 1 section section NOUN bracis-28409 39 2     SPACE bracis-28409 39 3 5 5 NUM bracis-28409 39 4 describes describe VERB bracis-28409 39 5 some some DET bracis-28409 39 6 methodological methodological ADJ bracis-28409 39 7 aspects aspect NOUN bracis-28409 39 8 of of ADP bracis-28409 39 9 the the DET bracis-28409 39 10 empirical empirical ADJ bracis-28409 39 11 analysis analysis NOUN bracis-28409 39 12 while while SCONJ bracis-28409 39 13 sect sect NOUN bracis-28409 39 14 . . PUNCT bracis-28409 39 15     SPACE bracis-28409 40 1 6 6 NUM bracis-28409 40 2 presents present VERB bracis-28409 40 3 the the DET bracis-28409 40 4 obtained obtain VERB bracis-28409 40 5 results result NOUN bracis-28409 40 6 . . PUNCT bracis-28409 41 1 finally finally ADV bracis-28409 41 2 , , PUNCT bracis-28409 41 3 sect sect NOUN bracis-28409 41 4 . . PUNCT bracis-28409 41 5     SPACE bracis-28409 42 1 7 7 NUM bracis-28409 42 2 highlights highlight NOUN bracis-28409 42 3 some some DET bracis-28409 42 4 final final ADJ bracis-28409 42 5 considerations consideration NOUN bracis-28409 42 6 and and CCONJ bracis-28409 42 7 future future ADJ bracis-28409 42 8 possibilities possibility NOUN bracis-28409 42 9 of of ADP bracis-28409 42 10 researches research NOUN bracis-28409 42 11 . . PUNCT bracis-28409 43 1 2 2 NUM bracis-28409 43 2 theoretical theoretical ADJ bracis-28409 43 3 aspects aspect NOUN bracis-28409 43 4 2.1 2.1 NUM bracis-28409 43 5 semi semi ADV bracis-28409 43 6 - - ADJ bracis-28409 43 7 supervised supervised ADJ bracis-28409 43 8 learning learning NOUN bracis-28409 43 9 according accord VERB bracis-28409 43 10 to to ADP bracis-28409 43 11 the the DET bracis-28409 43 12 tasks task NOUN bracis-28409 43 13 to to PART bracis-28409 43 14 which which PRON bracis-28409 43 15 they they PRON bracis-28409 43 16 are be AUX bracis-28409 43 17 applied apply VERB bracis-28409 43 18 , , PUNCT bracis-28409 43 19 the the DET bracis-28409 43 20 machine machine NOUN bracis-28409 43 21 learning learn VERB bracis-28409 43 22 algorithms algorithm NOUN bracis-28409 43 23 can can AUX bracis-28409 43 24 be be AUX bracis-28409 43 25 broadly broadly ADV bracis-28409 43 26 divided divide VERB bracis-28409 43 27 into into ADP bracis-28409 43 28 three three NUM bracis-28409 43 29 types type NOUN bracis-28409 43 30 : : PUNCT bracis-28409 43 31 1 1 X bracis-28409 43 32 ) ) PUNCT bracis-28409 43 33 supervised supervise VERB bracis-28409 43 34 learning learn VERB bracis-28409 43 35 predictive predictive ADJ bracis-28409 43 36 tasks task NOUN bracis-28409 43 37 ; ; PUNCT bracis-28409 43 38 2 2 X bracis-28409 43 39 ) ) PUNCT bracis-28409 43 40 unsupervised unsupervised ADJ bracis-28409 43 41 learning learn VERB bracis-28409 43 42 descriptive descriptive ADJ bracis-28409 43 43 tasks task NOUN bracis-28409 43 44 ; ; PUNCT bracis-28409 43 45 3 3 X bracis-28409 43 46 ) ) PUNCT bracis-28409 43 47 semi semi ADJ bracis-28409 43 48 - - ADJ bracis-28409 43 49 supervised supervised ADJ bracis-28409 43 50 learning learning NOUN bracis-28409 43 51 combining combine VERB bracis-28409 43 52 predictive predictive ADJ bracis-28409 43 53 and and CCONJ bracis-28409 43 54 descriptive descriptive ADJ bracis-28409 43 55 tasks task NOUN bracis-28409 43 56 . . PUNCT bracis-28409 44 1 in in ADP bracis-28409 44 2 other other ADJ bracis-28409 44 3 words word NOUN bracis-28409 44 4 , , PUNCT bracis-28409 44 5 supervised supervised ADJ bracis-28409 44 6 learning learning NOUN bracis-28409 44 7 requires require VERB bracis-28409 44 8 the the DET bracis-28409 44 9 training training NOUN bracis-28409 44 10 dataset dataset VERB bracis-28409 44 11 to to PART bracis-28409 44 12 have have VERB bracis-28409 44 13 input input NOUN bracis-28409 44 14 and and CCONJ bracis-28409 44 15 output output NOUN bracis-28409 44 16 attributes attribute NOUN bracis-28409 44 17 ( ( PUNCT bracis-28409 44 18 labels label NOUN bracis-28409 44 19 or or CCONJ bracis-28409 44 20 classes class NOUN bracis-28409 44 21 ) ) PUNCT bracis-28409 44 22 in in ADP bracis-28409 44 23 order order NOUN bracis-28409 44 24 to to PART bracis-28409 44 25 find find VERB bracis-28409 44 26 a a DET bracis-28409 44 27 model model NOUN bracis-28409 44 28 , , PUNCT bracis-28409 44 29 based base VERB bracis-28409 44 30 on on ADP bracis-28409 44 31 the the DET bracis-28409 44 32 data datum NOUN bracis-28409 44 33 used use VERB bracis-28409 44 34 for for ADP bracis-28409 44 35 training training NOUN bracis-28409 44 36 , , PUNCT bracis-28409 44 37 that that PRON bracis-28409 44 38 can can AUX bracis-28409 44 39 be be AUX bracis-28409 44 40 used use VERB bracis-28409 44 41 to to PART bracis-28409 44 42 predict predict VERB bracis-28409 44 43 a a DET bracis-28409 44 44 label label NOUN bracis-28409 44 45 or or CCONJ bracis-28409 44 46 class class NOUN bracis-28409 44 47 of of ADP bracis-28409 44 48 a a DET bracis-28409 44 49 given give VERB bracis-28409 44 50 instance instance NOUN bracis-28409 45 1 [ [ X bracis-28409 45 2 1 1 NUM bracis-28409 45 3 ] ] PUNCT bracis-28409 45 4 . . PUNCT bracis-28409 46 1 in in ADP bracis-28409 46 2 unsupervised unsupervised ADJ bracis-28409 46 3 learning learning NOUN bracis-28409 46 4 , , PUNCT bracis-28409 46 5 dataset dataset VERB bracis-28409 46 6 attributes attribute NOUN bracis-28409 46 7 are be AUX bracis-28409 46 8 not not PART bracis-28409 46 9 required require VERB bracis-28409 46 10 to to PART bracis-28409 46 11 have have VERB bracis-28409 46 12 an an DET bracis-28409 46 13 output output NOUN bracis-28409 46 14 or or CCONJ bracis-28409 46 15 label label NOUN bracis-28409 46 16 to to PART bracis-28409 46 17 shape shape VERB bracis-28409 46 18 the the DET bracis-28409 46 19 instances instance NOUN bracis-28409 46 20 ’ ' PUNCT bracis-28409 46 21 behavior behavior NOUN bracis-28409 46 22 pattern pattern NOUN bracis-28409 46 23 . . PUNCT bracis-28409 47 1 this this PRON bracis-28409 47 2 is be AUX bracis-28409 47 3 possible possible ADJ bracis-28409 47 4 due due ADP bracis-28409 47 5 to to ADP bracis-28409 47 6 the the DET bracis-28409 47 7 existence existence NOUN bracis-28409 47 8 of of ADP bracis-28409 47 9 tasks task NOUN bracis-28409 47 10 that that PRON bracis-28409 47 11 look look VERB bracis-28409 47 12 for for ADP bracis-28409 47 13 behavior behavior NOUN bracis-28409 47 14 patterns pattern NOUN bracis-28409 47 15 of of ADP bracis-28409 47 16 the the DET bracis-28409 47 17 instances instance NOUN bracis-28409 47 18 based base VERB bracis-28409 47 19 on on ADP bracis-28409 47 20 the the DET bracis-28409 47 21 input input NOUN bracis-28409 47 22 attributes attribute VERB bracis-28409 47 23 . . PUNCT bracis-28409 48 1 unsupervised unsupervised ADJ bracis-28409 48 2 or or CCONJ bracis-28409 48 3 descriptive descriptive ADJ bracis-28409 48 4 tasks task NOUN bracis-28409 48 5 are be AUX bracis-28409 48 6 more more ADV bracis-28409 48 7 related related ADJ bracis-28409 48 8 to to ADP bracis-28409 48 9 exploring explore VERB bracis-28409 48 10 the the DET bracis-28409 48 11 dataset dataset NOUN bracis-28409 48 12 , , PUNCT bracis-28409 48 13 with with ADP bracis-28409 48 14 the the DET bracis-28409 48 15 aim aim NOUN bracis-28409 48 16 of of ADP bracis-28409 48 17 obtaining obtain VERB bracis-28409 48 18 cluster cluster NOUN bracis-28409 48 19 patterns pattern NOUN bracis-28409 48 20 or or CCONJ bracis-28409 48 21 association association NOUN bracis-28409 48 22 rules rule NOUN bracis-28409 48 23 present present ADJ bracis-28409 48 24 in in ADP bracis-28409 48 25 the the DET bracis-28409 48 26 data datum NOUN bracis-28409 48 27 [ [ X bracis-28409 48 28 2 2 NUM bracis-28409 48 29 ] ] PUNCT bracis-28409 48 30 . . PUNCT bracis-28409 49 1 the the DET bracis-28409 49 2 existence existence NOUN bracis-28409 49 3 of of ADP bracis-28409 49 4 datasets dataset NOUN bracis-28409 49 5 that that PRON bracis-28409 49 6 have have VERB bracis-28409 49 7 a a DET bracis-28409 49 8 part part NOUN bracis-28409 49 9 of of ADP bracis-28409 49 10 the the DET bracis-28409 49 11 instances instance NOUN bracis-28409 49 12 with with ADP bracis-28409 49 13 labels label NOUN bracis-28409 49 14 and and CCONJ bracis-28409 49 15 another another DET bracis-28409 49 16 part part NOUN bracis-28409 49 17 without without ADP bracis-28409 49 18 assigned assign VERB bracis-28409 49 19 labels label NOUN bracis-28409 49 20 became become VERB bracis-28409 49 21 a a DET bracis-28409 49 22 motivation motivation NOUN bracis-28409 49 23 for for ADP bracis-28409 49 24 the the DET bracis-28409 49 25 creation creation NOUN bracis-28409 49 26 of of ADP bracis-28409 49 27 a a DET bracis-28409 49 28 new new ADJ bracis-28409 49 29 classification classification NOUN bracis-28409 49 30 model model NOUN bracis-28409 49 31 with with ADP bracis-28409 49 32 the the DET bracis-28409 49 33 objective objective NOUN bracis-28409 49 34 of of ADP bracis-28409 49 35 improving improve VERB bracis-28409 49 36 the the DET bracis-28409 49 37 performance performance NOUN bracis-28409 49 38 of of ADP bracis-28409 49 39 learning learn VERB bracis-28409 49 40 tasks task NOUN bracis-28409 49 41 benefiting benefit VERB bracis-28409 49 42 from from ADP bracis-28409 49 43 the the DET bracis-28409 49 44 strategy strategy NOUN bracis-28409 49 45 of of ADP bracis-28409 49 46 using use VERB bracis-28409 49 47 labeled label VERB bracis-28409 49 48 and and CCONJ bracis-28409 49 49 unlabeled unlabeled ADJ bracis-28409 49 50 data datum NOUN bracis-28409 49 51 , , PUNCT bracis-28409 49 52 known know VERB bracis-28409 49 53 as as ADP bracis-28409 49 54 semi semi ADJ bracis-28409 49 55 - - ADJ bracis-28409 49 56 supervised supervised ADJ bracis-28409 49 57 learning learning NOUN bracis-28409 49 58 [ [ X bracis-28409 49 59 10 10 NUM bracis-28409 49 60 ] ] PUNCT bracis-28409 49 61 . . PUNCT bracis-28409 50 1 following follow VERB bracis-28409 50 2 this this DET bracis-28409 50 3 concept concept NOUN bracis-28409 50 4 , , PUNCT bracis-28409 50 5 different different ADJ bracis-28409 50 6 algorithms algorithm NOUN bracis-28409 50 7 were be AUX bracis-28409 50 8 developed develop VERB bracis-28409 50 9 that that PRON bracis-28409 50 10 propose propose VERB bracis-28409 50 11 the the DET bracis-28409 50 12 classification classification NOUN bracis-28409 50 13 of of ADP bracis-28409 50 14 unlabeled unlabeled ADJ bracis-28409 50 15 instances instance NOUN bracis-28409 50 16 based base VERB bracis-28409 50 17 on on ADP bracis-28409 50 18 the the DET bracis-28409 50 19 ones one NOUN bracis-28409 50 20 with with ADP bracis-28409 50 21 a a DET bracis-28409 50 22 label label NOUN bracis-28409 50 23 , , PUNCT bracis-28409 50 24 including include VERB bracis-28409 50 25 self self NOUN bracis-28409 50 26 - - PUNCT bracis-28409 50 27 training training NOUN bracis-28409 50 28 , , PUNCT bracis-28409 50 29 which which PRON bracis-28409 50 30 will will AUX bracis-28409 50 31 be be AUX bracis-28409 50 32 used use VERB bracis-28409 50 33 in in ADP bracis-28409 50 34 this this DET bracis-28409 50 35 work work NOUN bracis-28409 50 36 . . PUNCT bracis-28409 51 1 the the DET bracis-28409 51 2 self self NOUN bracis-28409 51 3 - - PUNCT bracis-28409 51 4 training training NOUN bracis-28409 51 5 method method NOUN bracis-28409 51 6 is be AUX bracis-28409 51 7 characterized characterize VERB bracis-28409 51 8 by by ADP bracis-28409 51 9 using use VERB bracis-28409 51 10 its its PRON bracis-28409 51 11 own own ADJ bracis-28409 51 12 prediction prediction NOUN bracis-28409 51 13 results result NOUN bracis-28409 51 14 , , PUNCT bracis-28409 51 15 obtained obtain VERB bracis-28409 51 16 by by ADP bracis-28409 51 17 a a DET bracis-28409 51 18 single single ADJ bracis-28409 51 19 classifier classifier NOUN bracis-28409 51 20 , , PUNCT bracis-28409 51 21 and and CCONJ bracis-28409 51 22 applying apply VERB bracis-28409 51 23 a a DET bracis-28409 51 24 new new ADJ bracis-28409 51 25 training training NOUN bracis-28409 51 26 to to ADP bracis-28409 51 27 the the DET bracis-28409 51 28 model model NOUN bracis-28409 51 29 until until SCONJ bracis-28409 51 30 all all DET bracis-28409 51 31 instances instance NOUN bracis-28409 51 32 that that PRON bracis-28409 51 33 did do AUX bracis-28409 51 34 not not PART bracis-28409 51 35 have have VERB bracis-28409 51 36 labels label NOUN bracis-28409 51 37 are be AUX bracis-28409 51 38 properly properly ADV bracis-28409 51 39 labeled label VERB bracis-28409 51 40 [ [ PUNCT bracis-28409 51 41 11 11 NUM bracis-28409 51 42 ] ] PUNCT bracis-28409 51 43 . . PUNCT bracis-28409 52 1 the the DET bracis-28409 52 2 algorithm algorithm NOUN bracis-28409 52 3     SPACE bracis-28409 52 4 1 1 NUM bracis-28409 52 5 depicts depict VERB bracis-28409 52 6 the the DET bracis-28409 52 7 flow flow NOUN bracis-28409 52 8 of of ADP bracis-28409 52 9 instructions instruction NOUN bracis-28409 52 10 that that PRON bracis-28409 52 11 are be AUX bracis-28409 52 12 executed execute VERB bracis-28409 52 13 in in ADP bracis-28409 52 14 the the DET bracis-28409 52 15 self self NOUN bracis-28409 52 16 - - PUNCT bracis-28409 52 17 training training NOUN bracis-28409 52 18 process process NOUN bracis-28409 52 19 . . PUNCT bracis-28409 53 1 first first ADV bracis-28409 53 2 , , PUNCT bracis-28409 53 3 the the DET bracis-28409 53 4 database database NOUN bracis-28409 53 5 is be AUX bracis-28409 53 6 divided divide VERB bracis-28409 53 7 into into ADP bracis-28409 53 8 two two NUM bracis-28409 53 9 sets set NOUN bracis-28409 53 10 : : PUNCT bracis-28409 53 11 \(d_{l}\ \(d_{l}\ NOUN bracis-28409 53 12 ) ) PUNCT bracis-28409 53 13 labeled label VERB bracis-28409 53 14 instances instance NOUN bracis-28409 53 15 and and CCONJ bracis-28409 53 16 \(d_{u}\ \(d_{u}\ NOUN bracis-28409 53 17 ) ) PUNCT bracis-28409 53 18 unlabeled unlabele VERB bracis-28409 53 19 instances instance NOUN bracis-28409 53 20 ( ( PUNCT bracis-28409 53 21 line line NOUN bracis-28409 53 22 1 1 NUM bracis-28409 53 23 ) ) PUNCT bracis-28409 53 24 . . PUNCT bracis-28409 54 1 then then ADV bracis-28409 54 2 , , PUNCT bracis-28409 54 3 a a DET bracis-28409 54 4 classifier classifier NOUN bracis-28409 54 5 f f PROPN bracis-28409 54 6 is be AUX bracis-28409 54 7 generated generate VERB bracis-28409 54 8 through through ADP bracis-28409 54 9 \(d_{l}\ \(d_{l}\ PROPN bracis-28409 54 10 ) ) PUNCT bracis-28409 54 11 ( ( PUNCT bracis-28409 54 12 line line NOUN bracis-28409 54 13 3 3 NUM bracis-28409 54 14 ) ) PUNCT bracis-28409 54 15 . . PUNCT bracis-28409 55 1 after after ADP bracis-28409 55 2 assigning assign VERB bracis-28409 55 3 the the DET bracis-28409 55 4 new new ADJ bracis-28409 55 5 labels label NOUN bracis-28409 55 6 to to ADP bracis-28409 55 7 the the DET bracis-28409 55 8 unlabeled unlabeled ADJ bracis-28409 55 9 data datum NOUN bracis-28409 55 10 , , PUNCT bracis-28409 55 11 these these DET bracis-28409 55 12 instances instance NOUN bracis-28409 55 13 are be AUX bracis-28409 55 14 inserted insert VERB bracis-28409 55 15 into into ADP bracis-28409 55 16 the the DET bracis-28409 55 17 subset subset NOUN bracis-28409 55 18 s s X bracis-28409 55 19 which which PRON bracis-28409 55 20 will will AUX bracis-28409 55 21 be be AUX bracis-28409 55 22 removed remove VERB bracis-28409 55 23 from from ADP bracis-28409 55 24 \(d_{u}\ \(d_{u}\ NOUN bracis-28409 55 25 ) ) PUNCT bracis-28409 55 26 and and CCONJ bracis-28409 55 27 added add VERB bracis-28409 55 28 into into ADP bracis-28409 55 29 \(d_{l}\ \(d_{l}\ PROPN bracis-28409 55 30 ) ) PUNCT bracis-28409 55 31 ( ( PUNCT bracis-28409 55 32 lines line NOUN bracis-28409 55 33 4–6 4–6 PROPN bracis-28409 55 34 ) ) PUNCT bracis-28409 55 35 . . PUNCT bracis-28409 56 1 then then ADV bracis-28409 56 2 the the DET bracis-28409 56 3 process process NOUN bracis-28409 56 4 will will AUX bracis-28409 56 5 restart restart VERB bracis-28409 56 6 until until SCONJ bracis-28409 56 7 there there PRON bracis-28409 56 8 is be VERB bracis-28409 56 9 no no DET bracis-28409 56 10 more more ADV bracis-28409 56 11 unlabeled unlabeled ADJ bracis-28409 56 12 data datum NOUN bracis-28409 56 13 . . PUNCT bracis-28409 57 1 note note VERB bracis-28409 57 2 that that SCONJ bracis-28409 57 3 , , PUNCT bracis-28409 57 4 based base VERB bracis-28409 57 5 on on ADP bracis-28409 57 6 the the DET bracis-28409 57 7 results result NOUN bracis-28409 57 8 generated generate VERB bracis-28409 57 9 by by ADP bracis-28409 57 10 the the DET bracis-28409 57 11 classifier classifier NOUN bracis-28409 57 12 , , PUNCT bracis-28409 57 13 the the DET bracis-28409 57 14 model model NOUN bracis-28409 57 15 is be AUX bracis-28409 57 16 retrained retrain VERB bracis-28409 57 17 in in ADP bracis-28409 57 18 order order NOUN bracis-28409 57 19 to to PART bracis-28409 57 20 learn learn VERB bracis-28409 57 21 according accord VERB bracis-28409 57 22 to to ADP bracis-28409 57 23 its its PRON bracis-28409 57 24 assertiveness assertiveness NOUN bracis-28409 57 25 rate rate NOUN bracis-28409 57 26 . . PUNCT bracis-28409 58 1 algorithm algorithm NOUN bracis-28409 58 2 1 1 NUM bracis-28409 58 3 self self NOUN bracis-28409 58 4 - - PUNCT bracis-28409 58 5 training training NOUN bracis-28409 58 6 full full ADJ bracis-28409 58 7 size size NOUN bracis-28409 58 8 image image NOUN bracis-28409 58 9 in in ADP bracis-28409 58 10 [ [ X bracis-28409 58 11 12 12 NUM bracis-28409 58 12 ] ] PUNCT bracis-28409 58 13 , , PUNCT bracis-28409 58 14 the the DET bracis-28409 58 15 authors author NOUN bracis-28409 58 16 addressed address VERB bracis-28409 58 17 an an DET bracis-28409 58 18 exploratory exploratory ADJ bracis-28409 58 19 analysis analysis NOUN bracis-28409 58 20 of of ADP bracis-28409 58 21 the the DET bracis-28409 58 22 self self NOUN bracis-28409 58 23 - - PUNCT bracis-28409 58 24 training training NOUN bracis-28409 58 25 algorithm algorithm NOUN bracis-28409 58 26 , , PUNCT bracis-28409 58 27 regarding regard VERB bracis-28409 58 28 its its PRON bracis-28409 58 29 adaptation adaptation NOUN bracis-28409 58 30 by by ADP bracis-28409 58 31 inserting insert VERB bracis-28409 58 32 a a DET bracis-28409 58 33 new new ADJ bracis-28409 58 34 parameter parameter NOUN bracis-28409 58 35 , , PUNCT bracis-28409 58 36 called call VERB bracis-28409 58 37 confidence confidence NOUN bracis-28409 58 38 , , PUNCT bracis-28409 58 39 which which PRON bracis-28409 58 40 is be AUX bracis-28409 58 41 used use VERB bracis-28409 58 42 to to PART bracis-28409 58 43 select select VERB bracis-28409 58 44 and and CCONJ bracis-28409 58 45 to to PART bracis-28409 58 46 label label VERB bracis-28409 58 47 unlabeled unlabele VERB bracis-28409 58 48 instances instance NOUN bracis-28409 58 49 . . PUNCT bracis-28409 59 1 the the DET bracis-28409 59 2 main main ADJ bracis-28409 59 3 aim aim NOUN bracis-28409 59 4 is be AUX bracis-28409 59 5 to to PART bracis-28409 59 6 select select VERB bracis-28409 59 7 instances instance NOUN bracis-28409 59 8 that that PRON bracis-28409 59 9 have have VERB bracis-28409 59 10 a a DET bracis-28409 59 11 prediction prediction NOUN bracis-28409 59 12 higher high ADJ bracis-28409 59 13 than than ADP bracis-28409 59 14 a a DET bracis-28409 59 15 confidence confidence NOUN bracis-28409 59 16 threshold threshold NOUN bracis-28409 59 17 . . PUNCT bracis-28409 60 1 however however ADV bracis-28409 60 2 , , PUNCT bracis-28409 60 3 the the DET bracis-28409 60 4 confidence confidence NOUN bracis-28409 60 5 threshold threshold NOUN bracis-28409 60 6 is be AUX bracis-28409 60 7 initially initially ADV bracis-28409 60 8 defined define VERB bracis-28409 60 9 and and CCONJ bracis-28409 60 10 this this DET bracis-28409 60 11 value value NOUN bracis-28409 60 12 is be AUX bracis-28409 60 13 used use VERB bracis-28409 60 14 throughout throughout ADP bracis-28409 60 15 the the DET bracis-28409 60 16 whole whole ADJ bracis-28409 60 17 labeling labeling NOUN bracis-28409 60 18 process process NOUN bracis-28409 60 19 . . PUNCT bracis-28409 61 1 in in ADP bracis-28409 61 2 [ [ X bracis-28409 61 3 3 3 NUM bracis-28409 61 4 ] ] PUNCT bracis-28409 61 5 , , PUNCT bracis-28409 61 6 the the DET bracis-28409 61 7 authors author NOUN bracis-28409 61 8 proposed propose VERB bracis-28409 61 9 the the DET bracis-28409 61 10 flexcon flexcon NOUN bracis-28409 61 11 - - PUNCT bracis-28409 61 12 c c NOUN bracis-28409 61 13 method method NOUN bracis-28409 61 14 , , PUNCT bracis-28409 61 15 which which PRON bracis-28409 61 16 is be AUX bracis-28409 61 17 an an DET bracis-28409 61 18 extension extension NOUN bracis-28409 61 19 of of ADP bracis-28409 61 20 the the DET bracis-28409 61 21 work work NOUN bracis-28409 61 22 proposed propose VERB bracis-28409 61 23 in in ADP bracis-28409 61 24 [ [ X bracis-28409 61 25 12 12 NUM bracis-28409 61 26 ] ] PUNCT bracis-28409 61 27 , , PUNCT bracis-28409 61 28 which which PRON bracis-28409 61 29 applies apply VERB bracis-28409 61 30 an an DET bracis-28409 61 31 adaptive adaptive ADJ bracis-28409 61 32 ( ( PUNCT bracis-28409 61 33 or or CCONJ bracis-28409 61 34 flexible flexible ADJ bracis-28409 61 35 ) ) PUNCT bracis-28409 61 36 confidence confidence NOUN bracis-28409 61 37 threshold threshold NOUN bracis-28409 61 38 , , PUNCT bracis-28409 61 39 which which PRON bracis-28409 61 40 is be AUX bracis-28409 61 41 calculated calculate VERB bracis-28409 61 42 in in ADP bracis-28409 61 43 every every DET bracis-28409 61 44 iteration iteration NOUN bracis-28409 61 45 of of ADP bracis-28409 61 46 the the DET bracis-28409 61 47 semi semi ADJ bracis-28409 61 48 - - ADJ bracis-28409 61 49 supervised supervised ADJ bracis-28409 61 50 process process NOUN bracis-28409 61 51 . . PUNCT bracis-28409 62 1 the the DET bracis-28409 62 2 flexcon flexcon NOUN bracis-28409 62 3 - - PUNCT bracis-28409 62 4 c c NOUN bracis-28409 62 5 algorithm algorithm NOUN bracis-28409 62 6 is be AUX bracis-28409 62 7 presented present VERB bracis-28409 62 8 by by ADP bracis-28409 62 9 algorithm algorithm NOUN bracis-28409 62 10 2 2 NUM bracis-28409 62 11 . . PUNCT bracis-28409 62 12 algorithm algorithm NOUN bracis-28409 62 13 2 2 NUM bracis-28409 62 14 self self NOUN bracis-28409 62 15 - - PUNCT bracis-28409 62 16 training training NOUN bracis-28409 62 17 with with ADP bracis-28409 62 18 confidence confidence NOUN bracis-28409 62 19 adjustment adjustment NOUN bracis-28409 62 20 ( ( PUNCT bracis-28409 62 21 flexcon flexcon NOUN bracis-28409 62 22 - - PUNCT bracis-28409 62 23 c c NOUN bracis-28409 62 24 ) ) PUNCT bracis-28409 62 25 full full ADJ bracis-28409 62 26 size size NOUN bracis-28409 62 27 image image NOUN bracis-28409 62 28 in in ADP bracis-28409 62 29 algorithm algorithm NOUN bracis-28409 62 30 2 2 NUM bracis-28409 62 31 , , PUNCT bracis-28409 62 32 the the DET bracis-28409 62 33 lines line NOUN bracis-28409 62 34 marked mark VERB bracis-28409 62 35 in in ADP bracis-28409 62 36 blue blue ADJ bracis-28409 62 37 highlight highlight VERB bracis-28409 62 38 the the DET bracis-28409 62 39 main main ADJ bracis-28409 62 40 differences difference NOUN bracis-28409 62 41 between between ADP bracis-28409 62 42 this this DET bracis-28409 62 43 algorithm algorithm NOUN bracis-28409 62 44 and and CCONJ bracis-28409 62 45 the the DET bracis-28409 62 46 original original ADJ bracis-28409 62 47 self self NOUN bracis-28409 62 48 - - PUNCT bracis-28409 62 49 training training NOUN bracis-28409 62 50 algorithm algorithm NOUN bracis-28409 62 51 . . PUNCT bracis-28409 63 1 initially initially ADV bracis-28409 63 2 , , PUNCT bracis-28409 63 3 \(d_{l}\ \(d_{l}\ NOUN bracis-28409 63 4 ) ) PUNCT bracis-28409 63 5 represents represent VERB bracis-28409 63 6 the the DET bracis-28409 63 7 initially initially ADV bracis-28409 63 8 labeled label VERB bracis-28409 63 9 set set NOUN bracis-28409 63 10 , , PUNCT bracis-28409 63 11 while while SCONJ bracis-28409 63 12 \(d_{u}\ \(d_{u}\ NOUN bracis-28409 63 13 ) ) PUNCT bracis-28409 63 14 stands stand VERB bracis-28409 63 15 for for ADP bracis-28409 63 16 the the DET bracis-28409 63 17 set set NOUN bracis-28409 63 18 of of ADP bracis-28409 63 19 unlabeled unlabeled ADJ bracis-28409 63 20 instances instance NOUN bracis-28409 63 21 . . PUNCT bracis-28409 64 1 at at ADP bracis-28409 64 2 each each DET bracis-28409 64 3 iteration iteration NOUN bracis-28409 64 4 , , PUNCT bracis-28409 64 5 a a DET bracis-28409 64 6 classifier classifier NOUN bracis-28409 64 7 f f PROPN bracis-28409 64 8 is be AUX bracis-28409 64 9 trained train VERB bracis-28409 64 10 using use VERB bracis-28409 64 11 the the DET bracis-28409 64 12 set set NOUN bracis-28409 64 13 of of ADP bracis-28409 64 14 labeled label VERB bracis-28409 64 15 instances instance NOUN bracis-28409 64 16 \(d_{l}\ \(d_{l}\ NOUN bracis-28409 64 17 ) ) PUNCT bracis-28409 64 18 as as SCONJ bracis-28409 64 19 training training NOUN bracis-28409 64 20 set set NOUN bracis-28409 64 21 and and CCONJ bracis-28409 64 22 classifies classify VERB bracis-28409 64 23 the the DET bracis-28409 64 24 \(d_{u}\ \(d_{u}\ NOUN bracis-28409 64 25 ) ) PUNCT bracis-28409 64 26 instances instance NOUN bracis-28409 64 27 ( ( PUNCT bracis-28409 64 28 lines line NOUN bracis-28409 64 29 3 3 NUM bracis-28409 64 30 to to ADP bracis-28409 64 31 4 4 NUM bracis-28409 64 32 ) ) PUNCT bracis-28409 64 33 . . PUNCT bracis-28409 65 1 then then ADV bracis-28409 65 2 , , PUNCT bracis-28409 65 3 a a DET bracis-28409 65 4 new new ADJ bracis-28409 65 5 confidence confidence NOUN bracis-28409 65 6 threshold threshold NOUN bracis-28409 65 7 value value NOUN bracis-28409 65 8 is be AUX bracis-28409 65 9 calculated calculate VERB bracis-28409 65 10 ( ( PUNCT bracis-28409 65 11 line line NOUN bracis-28409 65 12 5 5 NUM bracis-28409 65 13 ) ) PUNCT bracis-28409 65 14 . . PUNCT bracis-28409 66 1 then then ADV bracis-28409 66 2 , , PUNCT bracis-28409 66 3 select select VERB bracis-28409 66 4 all all DET bracis-28409 66 5 unlabeled unlabele VERB bracis-28409 66 6 instances instance NOUN bracis-28409 66 7 whose whose DET bracis-28409 66 8 prediction prediction NOUN bracis-28409 66 9 is be AUX bracis-28409 66 10 higher high ADJ bracis-28409 66 11 than than ADP bracis-28409 66 12 or or CCONJ bracis-28409 66 13 equal equal ADJ bracis-28409 66 14 to to ADP bracis-28409 66 15 the the DET bracis-28409 66 16 confidence confidence NOUN bracis-28409 66 17 threshold threshold NOUN bracis-28409 66 18 . . PUNCT bracis-28409 67 1 these these DET bracis-28409 67 2 instances instance NOUN bracis-28409 67 3 are be AUX bracis-28409 67 4 then then ADV bracis-28409 67 5 labeled label VERB bracis-28409 67 6 ( ( PUNCT bracis-28409 67 7 line line NOUN bracis-28409 67 8 6 6 NUM bracis-28409 67 9 ) ) PUNCT bracis-28409 67 10 . . PUNCT bracis-28409 68 1 these these DET bracis-28409 68 2 instances instance NOUN bracis-28409 68 3 are be AUX bracis-28409 68 4 removed remove VERB bracis-28409 68 5 from from ADP bracis-28409 68 6 the the DET bracis-28409 68 7 \(d_{u}\ \(d_{u}\ NOUN bracis-28409 68 8 ) ) PUNCT bracis-28409 68 9 set set VERB bracis-28409 68 10 and and CCONJ bracis-28409 68 11 included include VERB bracis-28409 68 12 to to ADP bracis-28409 68 13 the the DET bracis-28409 68 14 \(d_{l}\ \(d_{l}\ NOUN bracis-28409 68 15 ) ) PUNCT bracis-28409 68 16 set set NOUN bracis-28409 68 17 . . PUNCT bracis-28409 69 1 this this DET bracis-28409 69 2 process process NOUN bracis-28409 69 3 is be AUX bracis-28409 69 4 then then ADV bracis-28409 69 5 repeated repeat VERB bracis-28409 69 6 until until ADP bracis-28409 69 7 the the DET bracis-28409 69 8 \(d_{u}\ \(d_{u}\ NOUN bracis-28409 69 9 ) ) PUNCT bracis-28409 69 10 set set NOUN bracis-28409 69 11 is be AUX bracis-28409 69 12 empty empty ADJ bracis-28409 69 13 . . PUNCT bracis-28409 70 1 in in ADP bracis-28409 70 2 [ [ X bracis-28409 70 3 3 3 NUM bracis-28409 70 4 ] ] PUNCT bracis-28409 70 5 , , PUNCT bracis-28409 70 6 three three NUM bracis-28409 70 7 ssl ssl ADJ bracis-28409 70 8 algorithms algorithm NOUN bracis-28409 70 9 are be AUX bracis-28409 70 10 proposed propose VERB bracis-28409 70 11 ( ( PUNCT bracis-28409 70 12 flexcon flexcon NOUN bracis-28409 70 13 - - PUNCT bracis-28409 70 14 g g NOUN bracis-28409 70 15 , , PUNCT bracis-28409 70 16 flexcon flexcon NOUN bracis-28409 70 17 , , PUNCT bracis-28409 70 18 and and CCONJ bracis-28409 70 19 flexcon flexcon NOUN bracis-28409 70 20 - - PUNCT bracis-28409 70 21 c c NOUN bracis-28409 70 22 ) ) PUNCT bracis-28409 70 23 . . PUNCT bracis-28409 71 1 all all DET bracis-28409 71 2 three three NUM bracis-28409 71 3 algorithms algorithm NOUN bracis-28409 71 4 use use VERB bracis-28409 71 5 adaptive adaptive ADJ bracis-28409 71 6 confidence confidence NOUN bracis-28409 71 7 threshold threshold NOUN bracis-28409 71 8 and and CCONJ bracis-28409 71 9 they they PRON bracis-28409 71 10 are be AUX bracis-28409 71 11 all all ADV bracis-28409 71 12 based base VERB bracis-28409 71 13 on on ADP bracis-28409 71 14 the the DET bracis-28409 71 15 selftraining selftraining NOUN bracis-28409 71 16 method method NOUN bracis-28409 71 17 . . PUNCT bracis-28409 72 1 in in ADP bracis-28409 72 2 the the DET bracis-28409 72 3 cited cite VERB bracis-28409 72 4 article article NOUN bracis-28409 72 5 , , PUNCT bracis-28409 72 6 the the DET bracis-28409 72 7 flexcon flexcon NOUN bracis-28409 72 8 - - PUNCT bracis-28409 72 9 c c NOUN bracis-28409 72 10 algorithm algorithm NOUN bracis-28409 72 11 adjusts adjust VERB bracis-28409 72 12 the the DET bracis-28409 72 13 confidence confidence NOUN bracis-28409 72 14 parameter parameter NOUN bracis-28409 72 15 at at ADP bracis-28409 72 16 each each DET bracis-28409 72 17 iteration iteration NOUN bracis-28409 72 18 based base VERB bracis-28409 72 19 solely solely ADV bracis-28409 72 20 on on ADP bracis-28409 72 21 the the DET bracis-28409 72 22 precision precision NOUN bracis-28409 72 23 of of ADP bracis-28409 72 24 a a DET bracis-28409 72 25 classifier classifier NOUN bracis-28409 72 26 built build VERB bracis-28409 72 27 the the DET bracis-28409 72 28 previous previous ADJ bracis-28409 72 29 iteration iteration NOUN bracis-28409 72 30 . . PUNCT bracis-28409 73 1 in in ADP bracis-28409 73 2 this this DET bracis-28409 73 3 method method NOUN bracis-28409 73 4 , , PUNCT bracis-28409 73 5 the the DET bracis-28409 73 6 adjustment adjustment NOUN bracis-28409 73 7 will will AUX bracis-28409 73 8 be be AUX bracis-28409 73 9 done do VERB bracis-28409 73 10 by by ADP bracis-28409 73 11 increasing increase VERB bracis-28409 73 12 or or CCONJ bracis-28409 73 13 decreasing decrease VERB bracis-28409 73 14 the the DET bracis-28409 73 15 confidence confidence NOUN bracis-28409 73 16 value value NOUN bracis-28409 73 17 based base VERB bracis-28409 73 18 on on ADP bracis-28409 73 19 a a DET bracis-28409 73 20 change change NOUN bracis-28409 73 21 rate rate NOUN bracis-28409 73 22 cr cr PROPN bracis-28409 73 23 . . PUNCT bracis-28409 73 24 equation equation NOUN bracis-28409 73 25     SPACE bracis-28409 73 26 1 1 NUM bracis-28409 73 27 presents present VERB bracis-28409 73 28 the the DET bracis-28409 73 29 confidence confidence NOUN bracis-28409 73 30 update update NOUN bracis-28409 73 31 equation equation NOUN bracis-28409 73 32 for for ADP bracis-28409 73 33 flexcon flexcon NOUN bracis-28409 73 34 - - PUNCT bracis-28409 73 35 c. c. NOUN bracis-28409 73 36 $ $ SYM bracis-28409 73 37 $ $ SYM bracis-28409 73 38 \begin{aligned \begin{aligne VERB bracis-28409 73 39 } } PUNCT bracis-28409 73 40 \textit{conf}(t_{i+1})={\left\ \textit{conf}(t_{i+1})={\left\ PROPN bracis-28409 73 41 { { PUNCT bracis-28409 73 42 \begin{array}{ll \begin{array}{ll PROPN bracis-28409 73 43 } } PUNCT bracis-28409 73 44 { { PUNCT bracis-28409 73 45 } } PUNCT bracis-28409 73 46 \textit{conf}(t_{i \textit{conf}(t_{i NOUN bracis-28409 73 47 } } PUNCT bracis-28409 73 48 ) ) PUNCT bracis-28409 74 1 \textit \textit NOUN bracis-28409 74 2 { { PUNCT bracis-28409 74 3 cr cr X bracis-28409 74 4 } } PUNCT bracis-28409 74 5 , , PUNCT bracis-28409 74 6 \textit{if \textit{if NOUN bracis-28409 74 7 } } PUNCT bracis-28409 74 8 \textit{acc \textit{acc PROPN bracis-28409 74 9 } } PUNCT bracis-28409 74 10 \ge \ge PROPN bracis-28409 74 11 \textit{mp \textit{mp NOUN bracis-28409 74 12 } } PUNCT bracis-28409 74 13 \textit \textit NOUN bracis-28409 74 14 { { PUNCT bracis-28409 74 15 + + NUM bracis-28409 74 16 e}\\ e}\\ NOUN bracis-28409 74 17 \textit{conf}(t_{i \textit{conf}(t_{i NOUN bracis-28409 74 18 } } PUNCT bracis-28409 74 19 ) ) PUNCT bracis-28409 74 20 , , PUNCT bracis-28409 74 21 \textit{if \textit{if NOUN bracis-28409 74 22 } } PUNCT bracis-28409 74 23 \textit{mp \textit{mp NOUN bracis-28409 74 24 } } PUNCT bracis-28409 74 25 \textit{e \textit{e NOUN bracis-28409 74 26 } } PUNCT bracis-28409 74 27 < < X bracis-28409 74 28 \textit{acc \textit{acc PROPN bracis-28409 74 29 } } PUNCT bracis-28409 74 30 < < X bracis-28409 74 31 \textit{mp \textit{mp NOUN bracis-28409 74 32 } } PUNCT bracis-28409 74 33 \textit \textit NOUN bracis-28409 74 34 { { PUNCT bracis-28409 74 35 + + NUM bracis-28409 74 36 e}\\ e}\\ NOUN bracis-28409 74 37 \textit{conf}(t_{i \textit{conf}(t_{i NOUN bracis-28409 74 38 } } PUNCT bracis-28409 74 39 ) ) PUNCT bracis-28409 75 1 \textit \textit NOUN bracis-28409 75 2 { { PUNCT bracis-28409 75 3 + + NUM bracis-28409 75 4 cr cr X bracis-28409 75 5 } } PUNCT bracis-28409 75 6 , , PUNCT bracis-28409 75 7 \textit{if \textit{if NOUN bracis-28409 75 8 } } PUNCT bracis-28409 75 9 \textit{acc \textit{acc PROPN bracis-28409 75 10 } } PUNCT bracis-28409 75 11 \le \le PROPN bracis-28409 75 12 \textit{mp \textit{mp NOUN bracis-28409 75 13 } } PUNCT bracis-28409 75 14 \textit{e \textit{e NOUN bracis-28409 75 15 } } PUNCT bracis-28409 75 16 \end{array}\right \end{array}\right PROPN bracis-28409 75 17 . . PUNCT bracis-28409 75 18 } } PUNCT bracis-28409 75 19 \end{aligned}$$ \end{aligned}$$ X bracis-28409 75 20 ( ( PUNCT bracis-28409 75 21 1 1 NUM bracis-28409 75 22 ) ) PUNCT bracis-28409 75 23 where where SCONJ bracis-28409 75 24 \(conf(t_{i+1})\ \(conf(t_{i+1})\ NOUN bracis-28409 75 25 ) ) PUNCT bracis-28409 75 26 is be AUX bracis-28409 75 27 the the DET bracis-28409 75 28 confidence confidence NOUN bracis-28409 75 29 value value NOUN bracis-28409 75 30 of of ADP bracis-28409 75 31 the the DET bracis-28409 75 32 current current ADJ bracis-28409 75 33 iteration iteration NOUN bracis-28409 75 34 , , PUNCT bracis-28409 75 35 mp mp PROPN bracis-28409 75 36 is be AUX bracis-28409 75 37 the the DET bracis-28409 75 38 minimum minimum ADJ bracis-28409 75 39 acceptable acceptable ADJ bracis-28409 75 40 precision precision NOUN bracis-28409 75 41 ; ; PUNCT bracis-28409 75 42 cr cr PROPN bracis-28409 75 43 is be AUX bracis-28409 75 44 the the DET bracis-28409 75 45 change change NOUN bracis-28409 75 46 rate rate NOUN bracis-28409 75 47 it it PRON bracis-28409 75 48 is be AUX bracis-28409 75 49 a a DET bracis-28409 75 50 user user NOUN bracis-28409 75 51 - - PUNCT bracis-28409 75 52 given give VERB bracis-28409 75 53 hyperparameter hyperparameter NOUN bracis-28409 75 54 , , PUNCT bracis-28409 75 55 acc acc PROPN bracis-28409 75 56 is be AUX bracis-28409 75 57 the the DET bracis-28409 75 58 accuracy accuracy NOUN bracis-28409 75 59 of of ADP bracis-28409 75 60 the the DET bracis-28409 75 61 classifier classifier NOUN bracis-28409 75 62 . . PUNCT bracis-28409 76 1 finally finally ADV bracis-28409 76 2 , , PUNCT bracis-28409 76 3 e e X bracis-28409 76 4 is be AUX bracis-28409 76 5 an an DET bracis-28409 76 6 acceptable acceptable ADJ bracis-28409 76 7 precision precision NOUN bracis-28409 76 8 variation variation NOUN bracis-28409 76 9 that that PRON bracis-28409 76 10 is be AUX bracis-28409 76 11 allowed allow VERB bracis-28409 76 12 in in ADP bracis-28409 76 13 order order NOUN bracis-28409 76 14 to to PART bracis-28409 76 15 define define VERB bracis-28409 76 16 an an DET bracis-28409 76 17 stabilization stabilization NOUN bracis-28409 76 18 in in ADP bracis-28409 76 19 precision precision NOUN bracis-28409 76 20 . . PUNCT bracis-28409 77 1 in in ADP bracis-28409 77 2 this this DET bracis-28409 77 3 equation equation NOUN bracis-28409 77 4 , , PUNCT bracis-28409 77 5 if if SCONJ bracis-28409 77 6 the the DET bracis-28409 77 7 accuracy accuracy NOUN bracis-28409 77 8 is be AUX bracis-28409 77 9 higher high ADJ bracis-28409 77 10 than than ADP bracis-28409 77 11 a a DET bracis-28409 77 12 minimum minimum ADJ bracis-28409 77 13 acceptable acceptable ADJ bracis-28409 77 14 precision precision NOUN bracis-28409 77 15 ( ( PUNCT bracis-28409 77 16 considering consider VERB bracis-28409 77 17 the the DET bracis-28409 77 18 acceptable acceptable ADJ bracis-28409 77 19 variation variation NOUN bracis-28409 77 20 e e NOUN bracis-28409 77 21 ) ) PUNCT bracis-28409 77 22 , , PUNCT bracis-28409 77 23 mp mp PROPN bracis-28409 77 24 , , PUNCT bracis-28409 77 25 the the DET bracis-28409 77 26 confidence confidence NOUN bracis-28409 77 27 threshold threshold NOUN bracis-28409 77 28 is be AUX bracis-28409 77 29 decreased decrease VERB bracis-28409 77 30 by by ADP bracis-28409 77 31 a a DET bracis-28409 77 32 change change NOUN bracis-28409 77 33 rate rate NOUN bracis-28409 77 34 cr cr PROPN bracis-28409 77 35 . . PUNCT bracis-28409 78 1 if if SCONJ bracis-28409 78 2 the the DET bracis-28409 78 3 accuracy accuracy NOUN bracis-28409 78 4 is be AUX bracis-28409 78 5 lower low ADJ bracis-28409 78 6 than than ADP bracis-28409 78 7 a a DET bracis-28409 78 8 minimum minimum ADJ bracis-28409 78 9 acceptable acceptable ADJ bracis-28409 78 10 precision precision NOUN bracis-28409 78 11 , , PUNCT bracis-28409 78 12 the the DET bracis-28409 78 13 confidence confidence NOUN bracis-28409 78 14 threshold threshold NOUN bracis-28409 78 15 is be AUX bracis-28409 78 16 increased increase VERB bracis-28409 78 17 by by ADP bracis-28409 78 18 a a DET bracis-28409 78 19 change change NOUN bracis-28409 78 20 rate rate NOUN bracis-28409 78 21 cr cr PROPN bracis-28409 78 22 . . PUNCT bracis-28409 79 1 otherwise otherwise ADV bracis-28409 79 2 , , PUNCT bracis-28409 79 3 the the DET bracis-28409 79 4 confidence confidence NOUN bracis-28409 79 5 threshold threshold NOUN bracis-28409 79 6 does do AUX bracis-28409 79 7 not not PART bracis-28409 79 8 change change VERB bracis-28409 79 9 . . PUNCT bracis-28409 80 1 this this DET bracis-28409 80 2 method method NOUN bracis-28409 80 3 was be AUX bracis-28409 80 4 further far ADV bracis-28409 80 5 divide divide NOUN bracis-28409 80 6 into into ADP bracis-28409 80 7 two two NUM bracis-28409 80 8 sub sub NOUN bracis-28409 80 9 - - NOUN bracis-28409 80 10 methods method NOUN bracis-28409 80 11 with with ADP bracis-28409 80 12 respect respect NOUN bracis-28409 80 13 to to ADP bracis-28409 80 14 the the DET bracis-28409 80 15 label label NOUN bracis-28409 80 16 definition definition NOUN bracis-28409 80 17 of of ADP bracis-28409 80 18 the the DET bracis-28409 80 19 unlabeled unlabele VERB bracis-28409 80 20 instances instance NOUN bracis-28409 80 21 , , PUNCT bracis-28409 80 22 flexcon flexcon NOUN bracis-28409 80 23 - - PUNCT bracis-28409 80 24 c1 c1 NOUN bracis-28409 80 25 and and CCONJ bracis-28409 80 26 flexcon flexcon NOUN bracis-28409 80 27 - - PUNCT bracis-28409 80 28 c2 c2 PROPN bracis-28409 80 29 . . PUNCT bracis-28409 81 1 flexcon flexcon PROPN bracis-28409 81 2 - - PUNCT bracis-28409 81 3 c1 c1 NOUN bracis-28409 81 4 uses use VERB bracis-28409 81 5 classifier classifier NOUN bracis-28409 81 6 ensembles ensemble NOUN bracis-28409 81 7 to to PART bracis-28409 81 8 define define VERB bracis-28409 81 9 the the DET bracis-28409 81 10 label label NOUN bracis-28409 81 11 of of ADP bracis-28409 81 12 each each DET bracis-28409 81 13 pattern pattern NOUN bracis-28409 81 14 , , PUNCT bracis-28409 81 15 while while SCONJ bracis-28409 81 16 flexcon flexcon NOUN bracis-28409 81 17 - - PUNCT bracis-28409 81 18 c2 c2 PROPN bracis-28409 81 19 uses use VERB bracis-28409 81 20 the the DET bracis-28409 81 21 label label NOUN bracis-28409 81 22 predicted predict VERB bracis-28409 81 23 by by ADP bracis-28409 81 24 the the DET bracis-28409 81 25 classifier classifier NOUN bracis-28409 81 26 generated generate VERB bracis-28409 81 27 in in ADP bracis-28409 81 28 the the DET bracis-28409 81 29 first first ADJ bracis-28409 81 30 iteration iteration NOUN bracis-28409 81 31 , , PUNCT bracis-28409 81 32 which which PRON bracis-28409 81 33 is be AUX bracis-28409 81 34 stored store VERB bracis-28409 81 35 by by ADP bracis-28409 81 36 the the DET bracis-28409 81 37 algorithm algorithm NOUN bracis-28409 81 38 . . PUNCT bracis-28409 82 1 in in ADP bracis-28409 82 2 flexcon flexcon NOUN bracis-28409 82 3 - - PUNCT bracis-28409 82 4 c1 c1 NOUN bracis-28409 82 5 , , PUNCT bracis-28409 82 6 the the DET bracis-28409 82 7 classifier classifier NOUN bracis-28409 82 8 ensemble ensemble ADJ bracis-28409 82 9 is be AUX bracis-28409 82 10 composed compose VERB bracis-28409 82 11 of of ADP bracis-28409 82 12 classifiers classifier NOUN bracis-28409 82 13 built build VERB bracis-28409 82 14 in in ADP bracis-28409 82 15 all all DET bracis-28409 82 16 iteration iteration NOUN bracis-28409 82 17 performed perform VERB bracis-28409 82 18 until until ADP bracis-28409 82 19 a a DET bracis-28409 82 20 current current ADJ bracis-28409 82 21 iteration iteration NOUN bracis-28409 82 22 . . PUNCT bracis-28409 83 1 in in ADP bracis-28409 83 2 addition addition NOUN bracis-28409 83 3 , , PUNCT bracis-28409 83 4 these these DET bracis-28409 83 5 individual individual ADJ bracis-28409 83 6 classifiers classifier NOUN bracis-28409 83 7 are be AUX bracis-28409 83 8 combined combine VERB bracis-28409 83 9 by by ADP bracis-28409 83 10 sum sum NOUN bracis-28409 83 11 and and CCONJ bracis-28409 83 12 majority majority NOUN bracis-28409 83 13 voting voting NOUN bracis-28409 83 14 , , PUNCT bracis-28409 83 15 leading lead VERB bracis-28409 83 16 to to ADP bracis-28409 83 17 two two NUM bracis-28409 83 18 versions version NOUN bracis-28409 83 19 of of ADP bracis-28409 83 20 this this DET bracis-28409 83 21 method method NOUN bracis-28409 83 22 , , PUNCT bracis-28409 83 23 flexcon flexcon NOUN bracis-28409 83 24 - - PUNCT bracis-28409 83 25 c1(v c1(v NOUN bracis-28409 83 26 ) ) PUNCT bracis-28409 83 27 and and CCONJ bracis-28409 83 28 flexcon flexcon NOUN bracis-28409 83 29 - - PUNCT bracis-28409 83 30 c1(s c1(s NOUN bracis-28409 83 31 ) ) PUNCT bracis-28409 83 32 . . PUNCT bracis-28409 84 1 2.2 2.2 NUM bracis-28409 84 2 classifier classifier NOUN bracis-28409 84 3 ensembles ensemble NOUN bracis-28409 84 4 as as ADV bracis-28409 84 5 previously previously ADV bracis-28409 84 6 mentioned mention VERB bracis-28409 84 7 , , PUNCT bracis-28409 84 8 the the DET bracis-28409 84 9 increased increase VERB bracis-28409 84 10 complexity complexity NOUN bracis-28409 84 11 and and CCONJ bracis-28409 84 12 wide wide ADJ bracis-28409 84 13 applicability applicability NOUN bracis-28409 84 14 of of ADP bracis-28409 84 15 classification classification NOUN bracis-28409 84 16 systems system NOUN bracis-28409 84 17 has have AUX bracis-28409 84 18 led lead VERB bracis-28409 84 19 to to ADP bracis-28409 84 20 exploring explore VERB bracis-28409 84 21 many many ADJ bracis-28409 84 22 approaches approach NOUN bracis-28409 84 23 and and CCONJ bracis-28409 84 24 methodologies methodology NOUN bracis-28409 84 25 . . PUNCT bracis-28409 85 1 nevertheless nevertheless ADV bracis-28409 85 2 , , PUNCT bracis-28409 85 3 there there PRON bracis-28409 85 4 is be VERB bracis-28409 85 5 a a DET bracis-28409 85 6 perception perception NOUN bracis-28409 85 7 that that SCONJ bracis-28409 85 8 no no DET bracis-28409 85 9 classifier classifier NOUN bracis-28409 85 10 is be AUX bracis-28409 85 11 considered consider VERB bracis-28409 85 12 completely completely ADV bracis-28409 85 13 satisfactory satisfactory ADJ bracis-28409 85 14 for for ADP bracis-28409 85 15 a a DET bracis-28409 85 16 particular particular ADJ bracis-28409 85 17 task task NOUN bracis-28409 85 18 ; ; PUNCT bracis-28409 85 19 therefore therefore ADV bracis-28409 85 20 , , PUNCT bracis-28409 85 21 the the DET bracis-28409 85 22 idea idea NOUN bracis-28409 85 23 of of ADP bracis-28409 85 24 combining combine VERB bracis-28409 85 25 different different ADJ bracis-28409 85 26 methods method NOUN bracis-28409 85 27 to to PART bracis-28409 85 28 improve improve VERB bracis-28409 85 29 performance performance NOUN bracis-28409 85 30 has have AUX bracis-28409 85 31 emerged emerge VERB bracis-28409 85 32 as as ADP bracis-28409 85 33 a a DET bracis-28409 85 34 very very ADV bracis-28409 85 35 promising promising ADJ bracis-28409 85 36 possibility possibility NOUN bracis-28409 85 37 [ [ X bracis-28409 85 38 13 13 NUM bracis-28409 85 39 ] ] PUNCT bracis-28409 85 40 . . PUNCT bracis-28409 86 1 this this DET bracis-28409 86 2 combination combination NOUN bracis-28409 86 3 is be AUX bracis-28409 86 4 called call VERB bracis-28409 86 5 classifier classifier ADJ bracis-28409 86 6 ensembles ensemble NOUN bracis-28409 86 7 , , PUNCT bracis-28409 86 8 also also ADV bracis-28409 86 9 known know VERB bracis-28409 86 10 as as ADP bracis-28409 86 11 multi multi ADJ bracis-28409 86 12 - - ADJ bracis-28409 86 13 classifier classifier ADJ bracis-28409 86 14 systems system NOUN bracis-28409 86 15 . . PUNCT bracis-28409 87 1 in in ADP bracis-28409 87 2 classification classification NOUN bracis-28409 87 3 tasks task NOUN bracis-28409 87 4 , , PUNCT bracis-28409 87 5 an an DET bracis-28409 87 6 ensemble ensemble ADJ bracis-28409 87 7 includes include VERB bracis-28409 87 8 several several ADJ bracis-28409 87 9 sub sub NOUN bracis-28409 87 10 - - NOUN bracis-28409 87 11 models model NOUN bracis-28409 87 12 called call VERB bracis-28409 87 13 base base NOUN bracis-28409 87 14 classifiers classifier NOUN bracis-28409 87 15 , , PUNCT bracis-28409 87 16 which which PRON bracis-28409 87 17 are be AUX bracis-28409 87 18 usually usually ADV bracis-28409 87 19 obtained obtain VERB bracis-28409 87 20 by by ADP bracis-28409 87 21 training train VERB bracis-28409 87 22 a a DET bracis-28409 87 23 basic basic ADJ bracis-28409 87 24 learning learning NOUN bracis-28409 87 25 algorithm algorithm NOUN bracis-28409 87 26 ( ( PUNCT bracis-28409 87 27 decision decision NOUN bracis-28409 87 28 tree tree NOUN bracis-28409 87 29 , , PUNCT bracis-28409 87 30 neural neural ADJ bracis-28409 87 31 network network NOUN bracis-28409 87 32 , , PUNCT bracis-28409 87 33 k k PROPN bracis-28409 87 34 nearest near ADJ bracis-28409 87 35 neighbors neighbor NOUN bracis-28409 87 36 , , PUNCT bracis-28409 87 37 among among ADP bracis-28409 87 38 others other NOUN bracis-28409 87 39 ) ) PUNCT bracis-28409 87 40 . . PUNCT bracis-28409 88 1 the the DET bracis-28409 88 2 ensembles ensemble NOUN bracis-28409 88 3 can can AUX bracis-28409 88 4 be be AUX bracis-28409 88 5 built build VERB bracis-28409 88 6 based base VERB bracis-28409 88 7 on on ADP bracis-28409 88 8 the the DET bracis-28409 88 9 same same ADJ bracis-28409 88 10 learning learning NOUN bracis-28409 88 11 algorithm algorithm NOUN bracis-28409 88 12 , , PUNCT bracis-28409 88 13 producing produce VERB bracis-28409 88 14 homogeneous homogeneous ADJ bracis-28409 88 15 ensembles ensemble NOUN bracis-28409 88 16 , , PUNCT bracis-28409 88 17 or or CCONJ bracis-28409 88 18 using use VERB bracis-28409 88 19 different different ADJ bracis-28409 88 20 algorithms algorithm NOUN bracis-28409 88 21 and and CCONJ bracis-28409 88 22 generating generate VERB bracis-28409 88 23 heterogeneous heterogeneous ADJ bracis-28409 88 24 ensembles ensemble NOUN bracis-28409 88 25     SPACE bracis-28409 88 26 [ [ X bracis-28409 88 27 14 14 NUM bracis-28409 88 28 ] ] PUNCT bracis-28409 88 29 . . PUNCT bracis-28409 89 1 the the DET bracis-28409 89 2 proposition proposition NOUN bracis-28409 89 3 of of ADP bracis-28409 89 4 classifier classifier ADJ bracis-28409 89 5 ensembles ensemble NOUN bracis-28409 89 6 is be AUX bracis-28409 89 7 to to PART bracis-28409 89 8 create create VERB bracis-28409 89 9 and and CCONJ bracis-28409 89 10 combine combine VERB bracis-28409 89 11 several several ADJ bracis-28409 89 12 inductive inductive ADJ bracis-28409 89 13 models model NOUN bracis-28409 89 14 for for ADP bracis-28409 89 15 the the DET bracis-28409 89 16 same same ADJ bracis-28409 89 17 domain domain NOUN bracis-28409 89 18 , , PUNCT bracis-28409 89 19 obtaining obtain VERB bracis-28409 89 20 better well ADJ bracis-28409 89 21 prediction prediction NOUN bracis-28409 89 22 quality quality NOUN bracis-28409 89 23 [ [ X bracis-28409 89 24 9 9 NUM bracis-28409 89 25 ] ] PUNCT bracis-28409 89 26 . . PUNCT bracis-28409 90 1 after after ADP bracis-28409 90 2 generating generate VERB bracis-28409 90 3 the the DET bracis-28409 90 4 base base NOUN bracis-28409 90 5 classifier classifier NOUN bracis-28409 90 6 set set NOUN bracis-28409 90 7 , , PUNCT bracis-28409 90 8 the the DET bracis-28409 90 9 next next ADJ bracis-28409 90 10 step step NOUN bracis-28409 90 11 is be AUX bracis-28409 90 12 the the DET bracis-28409 90 13 choice choice NOUN bracis-28409 90 14 for for SCONJ bracis-28409 90 15 the the DET bracis-28409 90 16 methods method NOUN bracis-28409 90 17 to to PART bracis-28409 90 18 combine combine VERB bracis-28409 90 19 their their PRON bracis-28409 90 20 outputs output NOUN bracis-28409 90 21 . . PUNCT bracis-28409 91 1 there there PRON bracis-28409 91 2 is be VERB bracis-28409 91 3 a a DET bracis-28409 91 4 vast vast ADJ bracis-28409 91 5 number number NOUN bracis-28409 91 6 of of ADP bracis-28409 91 7 methods method NOUN bracis-28409 91 8 for for ADP bracis-28409 91 9 combining combine VERB bracis-28409 91 10 classifiers classifier NOUN bracis-28409 91 11 in in ADP bracis-28409 91 12 the the DET bracis-28409 91 13 literature literature NOUN bracis-28409 91 14 . . PUNCT bracis-28409 92 1 in in ADP bracis-28409 92 2 this this DET bracis-28409 92 3 work work NOUN bracis-28409 92 4 , , PUNCT bracis-28409 92 5 we we PRON bracis-28409 92 6 will will AUX bracis-28409 92 7 be be AUX bracis-28409 92 8 used use VERB bracis-28409 92 9 voting vote VERB bracis-28409 92 10 . . PUNCT bracis-28409 93 1 the the DET bracis-28409 93 2 choice choice NOUN bracis-28409 93 3 for for ADP bracis-28409 93 4 this this DET bracis-28409 93 5 method method NOUN bracis-28409 93 6 was be AUX bracis-28409 93 7 made make VERB bracis-28409 93 8 due due ADJ bracis-28409 93 9 to to ADP bracis-28409 93 10 the the DET bracis-28409 93 11 use use NOUN bracis-28409 93 12 of of ADP bracis-28409 93 13 information information NOUN bracis-28409 93 14 from from ADP bracis-28409 93 15 all all DET bracis-28409 93 16 classifiers classifier NOUN bracis-28409 93 17 . . PUNCT bracis-28409 94 1 the the DET bracis-28409 94 2 voting voting NOUN bracis-28409 94 3 method method NOUN bracis-28409 94 4 , , PUNCT bracis-28409 94 5 often often ADV bracis-28409 94 6 used use VERB bracis-28409 94 7 to to PART bracis-28409 94 8 combine combine VERB bracis-28409 94 9 classifiers classifier NOUN bracis-28409 94 10 , , PUNCT bracis-28409 94 11 performs perform VERB bracis-28409 94 12 the the DET bracis-28409 94 13 combination combination NOUN bracis-28409 94 14 by by ADP bracis-28409 94 15 voting vote VERB bracis-28409 94 16 on on ADP bracis-28409 94 17 the the DET bracis-28409 94 18 results result NOUN bracis-28409 94 19 of of ADP bracis-28409 94 20 each each DET bracis-28409 94 21 classifier classifier NOUN bracis-28409 94 22 when when SCONJ bracis-28409 94 23 a a DET bracis-28409 94 24 new new ADJ bracis-28409 94 25 instance instance NOUN bracis-28409 94 26 is be AUX bracis-28409 94 27 presented present VERB bracis-28409 94 28 . . PUNCT bracis-28409 95 1 3 3 NUM bracis-28409 95 2 related relate VERB bracis-28409 95 3 work work NOUN bracis-28409 95 4 over over ADP bracis-28409 95 5 the the DET bracis-28409 95 6 years year NOUN bracis-28409 95 7 , , PUNCT bracis-28409 95 8 several several ADJ bracis-28409 95 9 works work NOUN bracis-28409 95 10 have have AUX bracis-28409 95 11 been be AUX bracis-28409 95 12 developed develop VERB bracis-28409 95 13 using use VERB bracis-28409 95 14 a a DET bracis-28409 95 15 confidence confidence NOUN bracis-28409 95 16 parameter parameter NOUN bracis-28409 95 17 in in ADP bracis-28409 95 18 semi semi ADJ bracis-28409 95 19 - - ADJ bracis-28409 95 20 supervised supervised ADJ bracis-28409 95 21 ml ml NOUN bracis-28409 95 22 algorithms algorithm NOUN bracis-28409 95 23 [ [ X bracis-28409 95 24 3 3 NUM bracis-28409 95 25 , , PUNCT bracis-28409 95 26 12 12 NUM bracis-28409 95 27 , , PUNCT bracis-28409 95 28 15,16,17,18 15,16,17,18 NUM bracis-28409 95 29 ] ] PUNCT bracis-28409 95 30 . . PUNCT bracis-28409 96 1 the the DET bracis-28409 96 2 goal goal NOUN bracis-28409 96 3 is be AUX bracis-28409 96 4 to to PART bracis-28409 96 5 investigate investigate VERB bracis-28409 96 6 the the DET bracis-28409 96 7 behavior behavior NOUN bracis-28409 96 8 of of ADP bracis-28409 96 9 these these DET bracis-28409 96 10 algorithms algorithm NOUN bracis-28409 96 11 , , PUNCT bracis-28409 96 12 including include VERB bracis-28409 96 13 this this DET bracis-28409 96 14 parameter parameter NOUN bracis-28409 96 15 as as ADP bracis-28409 96 16 a a DET bracis-28409 96 17 threshold threshold NOUN bracis-28409 96 18 for for ADP bracis-28409 96 19 selecting select VERB bracis-28409 96 20 new new ADJ bracis-28409 96 21 instances instance NOUN bracis-28409 96 22 for for ADP bracis-28409 96 23 the the DET bracis-28409 96 24 labeled label VERB bracis-28409 96 25 dataset dataset NOUN bracis-28409 96 26 . . PUNCT bracis-28409 97 1 on on ADP bracis-28409 97 2 the the DET bracis-28409 97 3 other other ADJ bracis-28409 97 4 hand hand NOUN bracis-28409 97 5 , , PUNCT bracis-28409 97 6 researchers researcher NOUN bracis-28409 97 7 have have AUX bracis-28409 97 8 explored explore VERB bracis-28409 97 9 the the DET bracis-28409 97 10 use use NOUN bracis-28409 97 11 of of ADP bracis-28409 97 12 classifier classifier ADJ bracis-28409 97 13 ensembles ensemble NOUN bracis-28409 97 14 to to PART bracis-28409 97 15 enhance enhance VERB bracis-28409 97 16 the the DET bracis-28409 97 17 performance performance NOUN bracis-28409 97 18 of of ADP bracis-28409 97 19 various various ADJ bracis-28409 97 20 machine machine NOUN bracis-28409 97 21 learning learn VERB bracis-28409 97 22 algorithms algorithm NOUN bracis-28409 97 23 [ [ X bracis-28409 97 24 4,5,6 4,5,6 NUM bracis-28409 97 25 , , PUNCT bracis-28409 97 26 8 8 NUM bracis-28409 97 27 ] ] PUNCT bracis-28409 97 28 . . PUNCT bracis-28409 98 1 the the DET bracis-28409 98 2 present present ADJ bracis-28409 98 3 work work NOUN bracis-28409 98 4 aims aim VERB bracis-28409 98 5 to to PART bracis-28409 98 6 merge merge VERB bracis-28409 98 7 these these DET bracis-28409 98 8 two two NUM bracis-28409 98 9 approaches approach NOUN bracis-28409 98 10 , , PUNCT bracis-28409 98 11 both both PRON bracis-28409 98 12 using use VERB bracis-28409 98 13 the the DET bracis-28409 98 14 confidence confidence NOUN bracis-28409 98 15 parameter parameter NOUN bracis-28409 98 16 as as ADV bracis-28409 98 17 well well ADV bracis-28409 98 18 as as ADP bracis-28409 98 19 classifier classifier NOUN bracis-28409 98 20 ensembles ensemble NOUN bracis-28409 98 21 , , PUNCT bracis-28409 98 22 as as ADP bracis-28409 98 23 a a DET bracis-28409 98 24 way way NOUN bracis-28409 98 25 to to PART bracis-28409 98 26 improve improve VERB bracis-28409 98 27 the the DET bracis-28409 98 28 performance performance NOUN bracis-28409 98 29 of of ADP bracis-28409 98 30 the the DET bracis-28409 98 31 semi semi ADJ bracis-28409 98 32 - - ADJ bracis-28409 98 33 supervised supervised ADJ bracis-28409 98 34 learning learning NOUN bracis-28409 98 35 algorithm algorithm NOUN bracis-28409 98 36 flexcon flexcon PROPN bracis-28409 98 37 - - PUNCT bracis-28409 98 38 c. c. NOUN bracis-28409 98 39 in in ADP bracis-28409 98 40 [ [ X bracis-28409 98 41 12 12 NUM bracis-28409 98 42 , , PUNCT bracis-28409 98 43 15 15 NUM bracis-28409 98 44 ] ] X bracis-28409 98 45 new new ADJ bracis-28409 98 46 versions version NOUN bracis-28409 98 47 of of ADP bracis-28409 98 48 the the DET bracis-28409 98 49 self self NOUN bracis-28409 98 50 - - PUNCT bracis-28409 98 51 training training NOUN bracis-28409 98 52 algorithm algorithm NOUN bracis-28409 98 53 were be AUX bracis-28409 98 54 proposed propose VERB bracis-28409 98 55 that that PRON bracis-28409 98 56 included include VERB bracis-28409 98 57 a a DET bracis-28409 98 58 fixed fix VERB bracis-28409 98 59 confidence confidence NOUN bracis-28409 98 60 parameter parameter NOUN bracis-28409 98 61 with with ADP bracis-28409 98 62 the the DET bracis-28409 98 63 objective objective NOUN bracis-28409 98 64 of of ADP bracis-28409 98 65 minimizing minimize VERB bracis-28409 98 66 inconsistencies inconsistency NOUN bracis-28409 98 67 in in ADP bracis-28409 98 68 the the DET bracis-28409 98 69 classification classification NOUN bracis-28409 98 70 choices choice NOUN bracis-28409 98 71 during during ADP bracis-28409 98 72 the the DET bracis-28409 98 73 labeling labeling NOUN bracis-28409 98 74 process process NOUN bracis-28409 98 75 . . PUNCT bracis-28409 99 1 in in ADP bracis-28409 99 2 [ [ X bracis-28409 99 3 3 3 NUM bracis-28409 99 4 , , PUNCT bracis-28409 99 5 16,17,18,19 16,17,18,19 NUM bracis-28409 99 6 ] ] PUNCT bracis-28409 99 7 algorithms algorithm NOUN bracis-28409 99 8 for for ADP bracis-28409 99 9 the the DET bracis-28409 99 10 automatic automatic ADJ bracis-28409 99 11 adjustment adjustment NOUN bracis-28409 99 12 of of ADP bracis-28409 99 13 the the DET bracis-28409 99 14 confidence confidence NOUN bracis-28409 99 15 threshold threshold NOUN bracis-28409 99 16 were be AUX bracis-28409 99 17 proposed propose VERB bracis-28409 99 18 as as ADP bracis-28409 99 19 an an DET bracis-28409 99 20 extension extension NOUN bracis-28409 99 21 of of ADP bracis-28409 99 22 those those DET bracis-28409 99 23 researches research NOUN bracis-28409 99 24 . . PUNCT bracis-28409 100 1 therefore therefore ADV bracis-28409 100 2 , , PUNCT bracis-28409 100 3 the the DET bracis-28409 100 4 focus focus NOUN bracis-28409 100 5 of of ADP bracis-28409 100 6 research research NOUN bracis-28409 100 7 development development NOUN bracis-28409 100 8 was be AUX bracis-28409 100 9 on on ADP bracis-28409 100 10 building build VERB bracis-28409 100 11 solutions solution NOUN bracis-28409 100 12 that that PRON bracis-28409 100 13 could could AUX bracis-28409 100 14 adjust adjust VERB bracis-28409 100 15 the the DET bracis-28409 100 16 confidence confidence NOUN bracis-28409 100 17 parameter parameter NOUN bracis-28409 100 18 during during ADP bracis-28409 100 19 the the DET bracis-28409 100 20 data datum NOUN bracis-28409 100 21 labeling labeling NOUN bracis-28409 100 22 process process NOUN bracis-28409 100 23 . . PUNCT bracis-28409 101 1 in in ADP bracis-28409 101 2 [ [ X bracis-28409 101 3 8 8 NUM bracis-28409 101 4 ] ] PUNCT bracis-28409 101 5 a a DET bracis-28409 101 6 study study NOUN bracis-28409 101 7 was be AUX bracis-28409 101 8 conducted conduct VERB bracis-28409 101 9 on on ADP bracis-28409 101 10 the the DET bracis-28409 101 11 existing exist VERB bracis-28409 101 12 risks risk NOUN bracis-28409 101 13 in in ADP bracis-28409 101 14 railway railway NOUN bracis-28409 101 15 safety safety NOUN bracis-28409 101 16 inspections inspection NOUN bracis-28409 101 17 . . PUNCT bracis-28409 102 1 these these DET bracis-28409 102 2 inspections inspection NOUN bracis-28409 102 3 are be AUX bracis-28409 102 4 reported report VERB bracis-28409 102 5 using use VERB bracis-28409 102 6 text text NOUN bracis-28409 102 7 , , PUNCT bracis-28409 102 8 which which PRON bracis-28409 102 9 generates generate VERB bracis-28409 102 10 a a DET bracis-28409 102 11 large large ADJ bracis-28409 102 12 amount amount NOUN bracis-28409 102 13 of of ADP bracis-28409 102 14 textual textual ADJ bracis-28409 102 15 data datum NOUN bracis-28409 102 16 . . PUNCT bracis-28409 103 1 as as ADP bracis-28409 103 2 motivation motivation NOUN bracis-28409 103 3 for for ADP bracis-28409 103 4 the the DET bracis-28409 103 5 authors author NOUN bracis-28409 103 6 , , PUNCT bracis-28409 103 7 a a DET bracis-28409 103 8 predictive predictive ADJ bracis-28409 103 9 model model NOUN bracis-28409 103 10 was be AUX bracis-28409 103 11 proposed propose VERB bracis-28409 103 12 that that SCONJ bracis-28409 103 13 used use VERB bracis-28409 103 14 classifier classifier NOUN bracis-28409 103 15 ensembles ensemble NOUN bracis-28409 103 16 to to PART bracis-28409 103 17 predict predict VERB bracis-28409 103 18 risk risk NOUN bracis-28409 103 19 situations situation NOUN bracis-28409 103 20 that that PRON bracis-28409 103 21 could could AUX bracis-28409 103 22 be be AUX bracis-28409 103 23 addressed address VERB bracis-28409 103 24 proactively proactively ADV bracis-28409 103 25 . . PUNCT bracis-28409 104 1 the the DET bracis-28409 104 2 chosen choose VERB bracis-28409 104 3 classification classification NOUN bracis-28409 104 4 model model NOUN bracis-28409 104 5 was be AUX bracis-28409 104 6 the the DET bracis-28409 104 7 decision decision NOUN bracis-28409 104 8 tree tree NOUN bracis-28409 104 9 , , PUNCT bracis-28409 104 10 which which PRON bracis-28409 104 11 was be AUX bracis-28409 104 12 applied apply VERB bracis-28409 104 13 to to ADP bracis-28409 104 14 an an DET bracis-28409 104 15 ensemble ensemble NOUN bracis-28409 104 16 of of ADP bracis-28409 104 17 bagging bag VERB bracis-28409 104 18 classifiers classifier NOUN bracis-28409 104 19 that that PRON bracis-28409 104 20 combined combine VERB bracis-28409 104 21 the the DET bracis-28409 104 22 results result NOUN bracis-28409 104 23 of of ADP bracis-28409 104 24 the the DET bracis-28409 104 25 classifiers classifier NOUN bracis-28409 104 26 by by ADP bracis-28409 104 27 voting vote VERB bracis-28409 104 28 . . PUNCT bracis-28409 105 1 similarly similarly ADV bracis-28409 105 2 , , PUNCT bracis-28409 105 3 [ [ X bracis-28409 105 4 5 5 NUM bracis-28409 105 5 ] ] PUNCT bracis-28409 105 6 used use VERB bracis-28409 105 7 different different ADJ bracis-28409 105 8 models model NOUN bracis-28409 105 9 of of ADP bracis-28409 105 10 classifiers classifier NOUN bracis-28409 105 11 in in ADP bracis-28409 105 12 an an DET bracis-28409 105 13 ensemble ensemble ADJ bracis-28409 105 14 to to PART bracis-28409 105 15 predict predict VERB bracis-28409 105 16 a a DET bracis-28409 105 17 credit credit NOUN bracis-28409 105 18 score score NOUN bracis-28409 105 19 for for ADP bracis-28409 105 20 customers customer NOUN bracis-28409 105 21 of of ADP bracis-28409 105 22 financial financial ADJ bracis-28409 105 23 organizations organization NOUN bracis-28409 105 24 and and CCONJ bracis-28409 105 25 banks bank NOUN bracis-28409 105 26 . . PUNCT bracis-28409 106 1 in in ADP bracis-28409 106 2 [ [ X bracis-28409 106 3 4 4 X bracis-28409 106 4 ] ] PUNCT bracis-28409 106 5 a a DET bracis-28409 106 6 new new ADJ bracis-28409 106 7 ensemble ensemble ADJ bracis-28409 106 8 learning learning NOUN bracis-28409 106 9 algorithm algorithm NOUN bracis-28409 106 10 named name VERB bracis-28409 106 11 e\(\_\)rhfs e\(\_\)rhfs PROPN bracis-28409 106 12 was be AUX bracis-28409 106 13 developed develop VERB bracis-28409 106 14 and and CCONJ bracis-28409 106 15 applied apply VERB bracis-28409 106 16 as as ADP bracis-28409 106 17 a a DET bracis-28409 106 18 tool tool NOUN bracis-28409 106 19 to to PART bracis-28409 106 20 predict predict VERB bracis-28409 106 21 software software NOUN bracis-28409 106 22 failure failure NOUN bracis-28409 106 23 . . PUNCT bracis-28409 107 1 the the DET bracis-28409 107 2 authors author NOUN bracis-28409 107 3 used use VERB bracis-28409 107 4 the the DET bracis-28409 107 5 voting voting NOUN bracis-28409 107 6 combination combination NOUN bracis-28409 107 7 method method NOUN bracis-28409 107 8 on on ADP bracis-28409 107 9 the the DET bracis-28409 107 10 base base NOUN bracis-28409 107 11 classifiers classifier NOUN bracis-28409 107 12 to to PART bracis-28409 107 13 obtain obtain VERB bracis-28409 107 14 the the DET bracis-28409 107 15 final final ADJ bracis-28409 107 16 ensemble ensemble ADJ bracis-28409 107 17 classifier classifier NOUN bracis-28409 107 18 for for ADP bracis-28409 107 19 software software NOUN bracis-28409 107 20 defect defect NOUN bracis-28409 107 21 prediction prediction NOUN bracis-28409 107 22 . . PUNCT bracis-28409 108 1 in in ADP bracis-28409 108 2 [ [ X bracis-28409 108 3 6 6 NUM bracis-28409 108 4 ] ] PUNCT bracis-28409 108 5 the the DET bracis-28409 108 6 proposal proposal NOUN bracis-28409 108 7 focused focus VERB bracis-28409 108 8 on on ADP bracis-28409 108 9 two two NUM bracis-28409 108 10 primary primary ADJ bracis-28409 108 11 factors factor NOUN bracis-28409 108 12 in in ADP bracis-28409 108 13 creating create VERB bracis-28409 108 14 a a DET bracis-28409 108 15 classifier classifier NOUN bracis-28409 108 16 ensemble ensemble ADJ bracis-28409 108 17 , , PUNCT bracis-28409 108 18 namely namely ADV bracis-28409 108 19 : : PUNCT bracis-28409 108 20 accuracy accuracy NOUN bracis-28409 108 21 and and CCONJ bracis-28409 108 22 diversity diversity NOUN bracis-28409 108 23 . . PUNCT bracis-28409 109 1 the the DET bracis-28409 109 2 paper paper NOUN bracis-28409 109 3 aimed aim VERB bracis-28409 109 4 to to PART bracis-28409 109 5 propose propose VERB bracis-28409 109 6 a a DET bracis-28409 109 7 new new ADJ bracis-28409 109 8 diversity diversity NOUN bracis-28409 109 9 measure measure NOUN bracis-28409 109 10 called call VERB bracis-28409 109 11 misclassification misclassification NOUN bracis-28409 109 12 diversity diversity NOUN bracis-28409 109 13 ( ( PUNCT bracis-28409 109 14 md md PROPN bracis-28409 109 15 ) ) PUNCT bracis-28409 109 16 and and CCONJ bracis-28409 109 17 an an DET bracis-28409 109 18 incremental incremental ADJ bracis-28409 109 19 layered layer VERB bracis-28409 109 20 classifier classifier NOUN bracis-28409 109 21 selection selection NOUN bracis-28409 109 22 approach approach NOUN bracis-28409 109 23 to to PART bracis-28409 109 24 build build VERB bracis-28409 109 25 the the DET bracis-28409 109 26 ensemble ensemble ADJ bracis-28409 109 27 . . PUNCT bracis-28409 110 1 4 4 NUM bracis-28409 110 2 the the DET bracis-28409 110 3 proposed propose VERB bracis-28409 110 4 method method NOUN bracis-28409 110 5 : : PUNCT bracis-28409 110 6 flexcon flexcon NOUN bracis-28409 110 7 - - PUNCT bracis-28409 110 8 ce ce PROPN bracis-28409 110 9 as as ADV bracis-28409 110 10 already already ADV bracis-28409 110 11 mentioned mention VERB bracis-28409 110 12 , , PUNCT bracis-28409 110 13 the the DET bracis-28409 110 14 purpose purpose NOUN bracis-28409 110 15 of of ADP bracis-28409 110 16 this this DET bracis-28409 110 17 paper paper NOUN bracis-28409 110 18 is be AUX bracis-28409 110 19 to to PART bracis-28409 110 20 propose propose VERB bracis-28409 110 21 a a DET bracis-28409 110 22 new new ADJ bracis-28409 110 23 flexcon flexcon NOUN bracis-28409 110 24 - - PUNCT bracis-28409 110 25 c c NOUN bracis-28409 110 26 variation variation NOUN bracis-28409 110 27 , , PUNCT bracis-28409 110 28 called call VERB bracis-28409 110 29 flexcon flexcon NOUN bracis-28409 110 30 - - PUNCT bracis-28409 110 31 ce ce PROPN bracis-28409 110 32 ( ( PUNCT bracis-28409 110 33 flexible flexible ADJ bracis-28409 110 34 confidence confidence NOUN bracis-28409 110 35 with with ADP bracis-28409 110 36 classifier classifier NOUN bracis-28409 110 37 ensemble ensemble ADJ bracis-28409 110 38 ) ) PUNCT bracis-28409 110 39 . . PUNCT bracis-28409 111 1 this this DET bracis-28409 111 2 variation variation NOUN bracis-28409 111 3 uses use VERB bracis-28409 111 4 an an DET bracis-28409 111 5 ensemble ensemble ADJ bracis-28409 111 6 - - PUNCT bracis-28409 111 7 based base VERB bracis-28409 111 8 confidence confidence NOUN bracis-28409 111 9 threshold threshold NOUN bracis-28409 111 10 . . PUNCT bracis-28409 112 1 therefore therefore ADV bracis-28409 112 2 , , PUNCT bracis-28409 112 3 in in ADP bracis-28409 112 4 order order NOUN bracis-28409 112 5 to to PART bracis-28409 112 6 develop develop VERB bracis-28409 112 7 the the DET bracis-28409 112 8 proposed propose VERB bracis-28409 112 9 method method NOUN bracis-28409 112 10 , , PUNCT bracis-28409 112 11 it it PRON bracis-28409 112 12 was be AUX bracis-28409 112 13 necessary necessary ADJ bracis-28409 112 14 to to PART bracis-28409 112 15 modify modify VERB bracis-28409 112 16 the the DET bracis-28409 112 17 original original ADJ bracis-28409 112 18 flexcon flexcon NOUN bracis-28409 112 19 - - PUNCT bracis-28409 112 20 c c NOUN bracis-28409 112 21 algorithm algorithm NOUN bracis-28409 112 22     SPACE bracis-28409 112 23 2 2 NUM bracis-28409 112 24 with with ADP bracis-28409 112 25 lines line NOUN bracis-28409 112 26 marked mark VERB bracis-28409 112 27 in in ADP bracis-28409 112 28 blue blue ADJ bracis-28409 112 29 highlight highlight NOUN bracis-28409 112 30 . . PUNCT bracis-28409 113 1 algorithm algorithm PROPN bracis-28409 113 2     SPACE bracis-28409 113 3 3 3 NUM bracis-28409 113 4 presents present VERB bracis-28409 113 5 the the DET bracis-28409 113 6 flexcon flexcon NOUN bracis-28409 113 7 - - PUNCT bracis-28409 113 8 ce ce NOUN bracis-28409 113 9 step step NOUN bracis-28409 113 10 - - PUNCT bracis-28409 113 11 by by ADP bracis-28409 113 12 - - PUNCT bracis-28409 113 13 step step NOUN bracis-28409 113 14 operation operation NOUN bracis-28409 113 15 . . PUNCT bracis-28409 114 1 algorithm algorithm NOUN bracis-28409 114 2 3 3 NUM bracis-28409 114 3 flexible flexible ADJ bracis-28409 114 4 confidence confidence NOUN bracis-28409 114 5 with with ADP bracis-28409 114 6 ensemble ensemble ADJ bracis-28409 114 7 ( ( PUNCT bracis-28409 114 8 flexcon flexcon NOUN bracis-28409 114 9 - - PUNCT bracis-28409 114 10 ce ce NOUN bracis-28409 114 11 ) ) PUNCT bracis-28409 114 12 full full ADJ bracis-28409 114 13 size size NOUN bracis-28409 114 14 image image NOUN bracis-28409 114 15 initially initially ADV bracis-28409 114 16 , , PUNCT bracis-28409 114 17 in in ADP bracis-28409 114 18 algorithm algorithm NOUN bracis-28409 114 19     SPACE bracis-28409 114 20 3 3 NUM bracis-28409 114 21 , , PUNCT bracis-28409 114 22 the the DET bracis-28409 114 23 database database NOUN bracis-28409 114 24 is be AUX bracis-28409 114 25 divided divide VERB bracis-28409 114 26 into into ADP bracis-28409 114 27 two two NUM bracis-28409 114 28 sets set NOUN bracis-28409 114 29 : : PUNCT bracis-28409 114 30 a a DET bracis-28409 114 31 set set NOUN bracis-28409 114 32 of of ADP bracis-28409 114 33 labeled label VERB bracis-28409 114 34 instances instance NOUN bracis-28409 114 35 \(d_{l}\ \(d_{l}\ NOUN bracis-28409 114 36 ) ) PUNCT bracis-28409 114 37 and and CCONJ bracis-28409 114 38 a a DET bracis-28409 114 39 set set NOUN bracis-28409 114 40 of of ADP bracis-28409 114 41 unlabeled unlabeled ADJ bracis-28409 114 42 instances instance NOUN bracis-28409 114 43 \(d_{u}\ \(d_{u}\ NOUN bracis-28409 114 44 ) ) PUNCT bracis-28409 114 45 . . PUNCT bracis-28409 115 1 additionally additionally ADV bracis-28409 115 2 , , PUNCT bracis-28409 115 3 a a DET bracis-28409 115 4 pool pool NOUN bracis-28409 115 5 of of ADP bracis-28409 115 6 classifiers classifier NOUN bracis-28409 115 7 \(f_{n}\ \(f_{n}\ X bracis-28409 115 8 ) ) PUNCT bracis-28409 115 9 is be AUX bracis-28409 115 10 available available ADJ bracis-28409 115 11 . . PUNCT bracis-28409 116 1 an an DET bracis-28409 116 2 ensemble ensemble ADJ bracis-28409 116 3 is be AUX bracis-28409 116 4 defined define VERB bracis-28409 116 5 as as ADP bracis-28409 116 6 empty empty ADJ bracis-28409 116 7 ( ( PUNCT bracis-28409 116 8 line line NOUN bracis-28409 116 9 1 1 NUM bracis-28409 116 10 ) ) PUNCT bracis-28409 116 11 . . PUNCT bracis-28409 117 1 then then ADV bracis-28409 117 2 , , PUNCT bracis-28409 117 3 for for ADP bracis-28409 117 4 each each DET bracis-28409 117 5 classifier classifier NOUN bracis-28409 117 6 f f PROPN bracis-28409 117 7 in in ADP bracis-28409 117 8 \(f_n\ \(f_n\ NOUN bracis-28409 117 9 ) ) PUNCT bracis-28409 117 10 , , PUNCT bracis-28409 117 11 they they PRON bracis-28409 117 12 are be AUX bracis-28409 117 13 trained train VERB bracis-28409 117 14 using use VERB bracis-28409 117 15 \(d_{l}\ \(d_{l}\ NOUN bracis-28409 117 16 ) ) PUNCT bracis-28409 117 17 as as ADP bracis-28409 117 18 training training NOUN bracis-28409 117 19 set set NOUN bracis-28409 117 20 . . PUNCT bracis-28409 118 1 after after ADP bracis-28409 118 2 that that PRON bracis-28409 118 3 , , PUNCT bracis-28409 118 4 the the DET bracis-28409 118 5 trained train VERB bracis-28409 118 6 classifier classifier NOUN bracis-28409 118 7 f f PROPN bracis-28409 118 8 is be AUX bracis-28409 118 9 added add VERB bracis-28409 118 10 to to ADP bracis-28409 118 11 the the DET bracis-28409 118 12 ensemble ensemble ADJ bracis-28409 118 13 e e NOUN bracis-28409 118 14 ( ( PUNCT bracis-28409 118 15 lines line NOUN bracis-28409 118 16 2 2 NUM bracis-28409 118 17 to to ADP bracis-28409 118 18 7 7 NUM bracis-28409 118 19 ) ) PUNCT bracis-28409 118 20 . . PUNCT bracis-28409 119 1 finally finally ADV bracis-28409 119 2 , , PUNCT bracis-28409 119 3 the the DET bracis-28409 119 4 ensemble ensemble ADJ bracis-28409 119 5 e e NOUN bracis-28409 119 6 will will AUX bracis-28409 119 7 be be AUX bracis-28409 119 8 applied apply VERB bracis-28409 119 9 to to ADP bracis-28409 119 10 \(d_{u}\ \(d_{u}\ NOUN bracis-28409 119 11 ) ) PUNCT bracis-28409 119 12 in in ADP bracis-28409 119 13 order order NOUN bracis-28409 119 14 to to PART bracis-28409 119 15 calculate calculate VERB bracis-28409 119 16 the the DET bracis-28409 119 17 effectiveness effectiveness NOUN bracis-28409 119 18 generated generate VERB bracis-28409 119 19 by by ADP bracis-28409 119 20 e e PROPN bracis-28409 119 21 ( ( PUNCT bracis-28409 119 22 line line NOUN bracis-28409 119 23 8) 8) NUM bracis-28409 119 24 . . PUNCT bracis-28409 120 1 then then ADV bracis-28409 120 2 , , PUNCT bracis-28409 120 3 the the DET bracis-28409 120 4 new new ADJ bracis-28409 120 5 confidence confidence NOUN bracis-28409 120 6 threshold threshold NOUN bracis-28409 120 7 is be AUX bracis-28409 120 8 calculated calculate VERB bracis-28409 120 9 ( ( PUNCT bracis-28409 120 10 line line NOUN bracis-28409 120 11 9 9 NUM bracis-28409 120 12 ) ) PUNCT bracis-28409 120 13 , , PUNCT bracis-28409 120 14 the the DET bracis-28409 120 15 unlabeled unlabele VERB bracis-28409 120 16 instances instance NOUN bracis-28409 120 17 are be AUX bracis-28409 120 18 selected select VERB bracis-28409 120 19 , , PUNCT bracis-28409 120 20 labeled label VERB bracis-28409 120 21 , , PUNCT bracis-28409 120 22 removed remove VERB bracis-28409 120 23 from from ADP bracis-28409 120 24 the the DET bracis-28409 120 25 unlabeled unlabele VERB bracis-28409 120 26 set set VERB bracis-28409 120 27 \(d_{u}\ \(d_{u}\ NOUN bracis-28409 120 28 ) ) PUNCT bracis-28409 120 29 and and CCONJ bracis-28409 120 30 included include VERB bracis-28409 120 31 in in ADP bracis-28409 120 32 the the DET bracis-28409 120 33 labeled label VERB bracis-28409 120 34 set set NOUN bracis-28409 120 35 \(d_{l}\ \(d_{l}\ NOUN bracis-28409 120 36 ) ) PUNCT bracis-28409 120 37 ( ( PUNCT bracis-28409 120 38 lines line NOUN bracis-28409 120 39 9 9 NUM bracis-28409 120 40 to to ADP bracis-28409 120 41 12 12 NUM bracis-28409 120 42 ) ) PUNCT bracis-28409 120 43 . . PUNCT bracis-28409 121 1 it it PRON bracis-28409 121 2 is be AUX bracis-28409 121 3 important important ADJ bracis-28409 121 4 to to PART bracis-28409 121 5 emphatize emphatize VERB bracis-28409 121 6 that that SCONJ bracis-28409 121 7 the the DET bracis-28409 121 8 threshold threshold NOUN bracis-28409 121 9 is be AUX bracis-28409 121 10 used use VERB bracis-28409 121 11 to to PART bracis-28409 121 12 select select VERB bracis-28409 121 13 unlabeled unlabeled ADJ bracis-28409 121 14 instances instance NOUN bracis-28409 121 15 that that PRON bracis-28409 121 16 will will AUX bracis-28409 121 17 be be AUX bracis-28409 121 18 labeled label VERB bracis-28409 121 19 by by ADP bracis-28409 121 20 the the DET bracis-28409 121 21 voting voting NOUN bracis-28409 121 22 process process NOUN bracis-28409 121 23 of of ADP bracis-28409 121 24 the the DET bracis-28409 121 25 ensemble ensemble NOUN bracis-28409 121 26 . . PUNCT bracis-28409 122 1 it it PRON bracis-28409 122 2 is be AUX bracis-28409 122 3 important important ADJ bracis-28409 122 4 to to PART bracis-28409 122 5 emphasize emphasize VERB bracis-28409 122 6 that that SCONJ bracis-28409 122 7 the the DET bracis-28409 122 8 calculation calculation NOUN bracis-28409 122 9 of of ADP bracis-28409 122 10 the the DET bracis-28409 122 11 new new ADJ bracis-28409 122 12 confidence confidence NOUN bracis-28409 122 13 threshold threshold NOUN bracis-28409 122 14 is be AUX bracis-28409 122 15 made make VERB bracis-28409 122 16 using use VERB bracis-28409 122 17 eq eq ADP bracis-28409 122 18 . . PUNCT bracis-28409 122 19     SPACE bracis-28409 122 20 1 1 X bracis-28409 122 21 . . PUNCT bracis-28409 123 1 the the DET bracis-28409 123 2 main main ADJ bracis-28409 123 3 difference difference NOUN bracis-28409 123 4 is be AUX bracis-28409 123 5 that that SCONJ bracis-28409 123 6 the the DET bracis-28409 123 7 mp mp NOUN bracis-28409 123 8 value value NOUN bracis-28409 123 9 is be AUX bracis-28409 123 10 related relate VERB bracis-28409 123 11 to to ADP bracis-28409 123 12 the the DET bracis-28409 123 13 ensemble ensemble ADJ bracis-28409 123 14 accuracy accuracy NOUN bracis-28409 123 15 ( ( PUNCT bracis-28409 123 16 flexcon flexcon NOUN bracis-28409 123 17 - - PUNCT bracis-28409 123 18 ce ce NOUN bracis-28409 123 19 ) ) PUNCT bracis-28409 123 20 instead instead ADV bracis-28409 123 21 of of ADP bracis-28409 123 22 the the DET bracis-28409 123 23 classifier classifier NOUN bracis-28409 123 24 accuracy accuracy NOUN bracis-28409 123 25 ( ( PUNCT bracis-28409 123 26 flexcon flexcon NOUN bracis-28409 123 27 - - PUNCT bracis-28409 123 28 c c NOUN bracis-28409 123 29 ) ) PUNCT bracis-28409 123 30 . . PUNCT bracis-28409 124 1 as as SCONJ bracis-28409 124 2 previously previously ADV bracis-28409 124 3 described describe VERB bracis-28409 124 4 , , PUNCT bracis-28409 124 5 two two NUM bracis-28409 124 6 flexcon flexcon NOUN bracis-28409 124 7 - - PUNCT bracis-28409 124 8 ce ce NOUN bracis-28409 124 9 versions version NOUN bracis-28409 124 10 are be AUX bracis-28409 124 11 proposed propose VERB bracis-28409 124 12 . . PUNCT bracis-28409 125 1 in in ADP bracis-28409 125 2 the the DET bracis-28409 125 3 first first ADJ bracis-28409 125 4 one one NUM bracis-28409 125 5 , , PUNCT bracis-28409 125 6 the the DET bracis-28409 125 7 classifier classifier NOUN bracis-28409 125 8 ensemble ensemble ADJ bracis-28409 125 9 is be AUX bracis-28409 125 10 constructed construct VERB bracis-28409 125 11 heterogeneously heterogeneously ADV bracis-28409 125 12 , , PUNCT bracis-28409 125 13 using use VERB bracis-28409 125 14 more more ADJ bracis-28409 125 15 than than ADP bracis-28409 125 16 one one NUM bracis-28409 125 17 type type NOUN bracis-28409 125 18 of of ADP bracis-28409 125 19 classification classification NOUN bracis-28409 125 20 method method NOUN bracis-28409 125 21 . . PUNCT bracis-28409 126 1 in in ADP bracis-28409 126 2 the the DET bracis-28409 126 3 second second ADJ bracis-28409 126 4 version version NOUN bracis-28409 126 5 , , PUNCT bracis-28409 126 6 the the DET bracis-28409 126 7 classifier classifier NOUN bracis-28409 126 8 ensemble ensemble ADJ bracis-28409 126 9 is be AUX bracis-28409 126 10 built build VERB bracis-28409 126 11 homogeneously homogeneously ADV bracis-28409 126 12 , , PUNCT bracis-28409 126 13 when when SCONJ bracis-28409 126 14 using use VERB bracis-28409 126 15 only only ADV bracis-28409 126 16 one one NUM bracis-28409 126 17 type type NOUN bracis-28409 126 18 of of ADP bracis-28409 126 19 method method NOUN bracis-28409 126 20 . . PUNCT bracis-28409 127 1 in in ADP bracis-28409 127 2 this this DET bracis-28409 127 3 work work NOUN bracis-28409 127 4 , , PUNCT bracis-28409 127 5 both both CCONJ bracis-28409 127 6 heterogeneous heterogeneous ADJ bracis-28409 127 7 and and CCONJ bracis-28409 127 8 homogeneous homogeneous ADJ bracis-28409 127 9 ensemble ensemble ADJ bracis-28409 127 10 approaches approach NOUN bracis-28409 127 11 will will AUX bracis-28409 127 12 be be AUX bracis-28409 127 13 assessed assess VERB bracis-28409 127 14 . . PUNCT bracis-28409 128 1 5 5 NUM bracis-28409 128 2 experimental experimental ADJ bracis-28409 128 3 methodology methodology NOUN bracis-28409 128 4 in in ADP bracis-28409 128 5 order order NOUN bracis-28409 128 6 to to PART bracis-28409 128 7 assess assess VERB bracis-28409 128 8 the the DET bracis-28409 128 9 feasibility feasibility NOUN bracis-28409 128 10 of of ADP bracis-28409 128 11 the the DET bracis-28409 128 12 proposed propose VERB bracis-28409 128 13 method method NOUN bracis-28409 128 14 , , PUNCT bracis-28409 128 15 an an DET bracis-28409 128 16 empirical empirical ADJ bracis-28409 128 17 analysis analysis NOUN bracis-28409 128 18 will will AUX bracis-28409 128 19 be be AUX bracis-28409 128 20 carried carry VERB bracis-28409 128 21 out out ADP bracis-28409 128 22 . . PUNCT bracis-28409 129 1 the the DET bracis-28409 129 2 next next ADJ bracis-28409 129 3 subsections subsection NOUN bracis-28409 129 4 will will AUX bracis-28409 129 5 describe describe VERB bracis-28409 129 6 the the DET bracis-28409 129 7 main main ADJ bracis-28409 129 8 aspects aspect NOUN bracis-28409 129 9 of of ADP bracis-28409 129 10 the the DET bracis-28409 129 11 experimental experimental ADJ bracis-28409 129 12 methodology methodology NOUN bracis-28409 129 13 . . PUNCT bracis-28409 130 1 5.1 5.1 NUM bracis-28409 130 2 datasets dataset NOUN bracis-28409 130 3 in in ADP bracis-28409 130 4 this this DET bracis-28409 130 5 empirical empirical ADJ bracis-28409 130 6 analysis analysis NOUN bracis-28409 130 7 , , PUNCT bracis-28409 130 8 the the DET bracis-28409 130 9 flexcon flexcon NOUN bracis-28409 130 10 - - PUNCT bracis-28409 130 11 ce ce PROPN bracis-28409 130 12 method method NOUN bracis-28409 130 13 is be AUX bracis-28409 130 14 applied apply VERB bracis-28409 130 15 to to ADP bracis-28409 130 16 20 20 NUM bracis-28409 130 17 different different ADJ bracis-28409 130 18 databases database NOUN bracis-28409 130 19 . . PUNCT bracis-28409 131 1 these these DET bracis-28409 131 2 datasets dataset NOUN bracis-28409 131 3 are be AUX bracis-28409 131 4 available available ADJ bracis-28409 131 5 in in ADP bracis-28409 131 6 platforms platform NOUN bracis-28409 131 7 and and CCONJ bracis-28409 131 8 repositories repository NOUN bracis-28409 131 9 that that PRON bracis-28409 131 10 maintain maintain VERB bracis-28409 131 11 various various ADJ bracis-28409 131 12 experimental experimental ADJ bracis-28409 131 13 data datum NOUN bracis-28409 131 14 , , PUNCT bracis-28409 131 15 such such ADJ bracis-28409 131 16 as as ADP bracis-28409 131 17 github github NOUN bracis-28409 131 18 [ [ X bracis-28409 131 19 20 20 NUM bracis-28409 131 20 ] ] PUNCT bracis-28409 131 21 , , PUNCT bracis-28409 131 22 uci uci PROPN bracis-28409 131 23 machine machine NOUN bracis-28409 131 24 learning learn VERB bracis-28409 131 25 [ [ X bracis-28409 131 26 21 21 NUM bracis-28409 131 27 ] ] PUNCT bracis-28409 131 28 and and CCONJ bracis-28409 131 29 kaggle kaggle VERB bracis-28409 131 30 datasets dataset NOUN bracis-28409 132 1 [ [ X bracis-28409 132 2 22 22 NUM bracis-28409 132 3 ] ] PUNCT bracis-28409 132 4 . . PUNCT bracis-28409 133 1 table table NOUN bracis-28409 133 2     SPACE bracis-28409 133 3 1 1 NUM bracis-28409 133 4 describes describe VERB bracis-28409 133 5 the the DET bracis-28409 133 6 characteristics characteristic NOUN bracis-28409 133 7 of of ADP bracis-28409 133 8 all all DET bracis-28409 133 9 datasets dataset NOUN bracis-28409 133 10 , , PUNCT bracis-28409 133 11 regarding regard VERB bracis-28409 133 12 the the DET bracis-28409 133 13 number number NOUN bracis-28409 133 14 of of ADP bracis-28409 133 15 attributes attribute NOUN bracis-28409 133 16 , , PUNCT bracis-28409 133 17 instances instance NOUN bracis-28409 133 18 , , PUNCT bracis-28409 133 19 number number NOUN bracis-28409 133 20 of of ADP bracis-28409 133 21 classes class NOUN bracis-28409 133 22 , , PUNCT bracis-28409 133 23 data datum NOUN bracis-28409 133 24 types type NOUN bracis-28409 133 25 ( ( PUNCT bracis-28409 133 26 categorical categorical ADJ bracis-28409 133 27 c c NOUN bracis-28409 133 28 , , PUNCT bracis-28409 133 29 real real ADJ bracis-28409 133 30 r r NOUN bracis-28409 133 31 and and CCONJ bracis-28409 133 32 integer integer NOUN bracis-28409 133 33 i i PROPN bracis-28409 133 34 ) ) PUNCT bracis-28409 133 35 , , PUNCT bracis-28409 133 36 and and CCONJ bracis-28409 133 37 data datum NOUN bracis-28409 133 38 distribution distribution NOUN bracis-28409 133 39 ( ( PUNCT bracis-28409 133 40 balanced balanced ADJ bracis-28409 133 41 b b NOUN bracis-28409 133 42 and and CCONJ bracis-28409 133 43 unbalanced unbalanced ADJ bracis-28409 133 44 u u NOUN bracis-28409 133 45 ) ) PUNCT bracis-28409 133 46 . . PUNCT bracis-28409 134 1 table table NOUN bracis-28409 134 2 1 1 NUM bracis-28409 134 3 . . PUNCT bracis-28409 135 1 datasets dataset NOUN bracis-28409 135 2 characteristicsfull characteristicsfull ADJ bracis-28409 135 3 size size NOUN bracis-28409 135 4 table table NOUN bracis-28409 135 5 5.2 5.2 NUM bracis-28409 135 6 methods method NOUN bracis-28409 135 7 and and CCONJ bracis-28409 135 8     SPACE bracis-28409 135 9 materials material NOUN bracis-28409 135 10 all all DET bracis-28409 135 11 method method NOUN bracis-28409 135 12 and and CCONJ bracis-28409 135 13 learning learn VERB bracis-28409 135 14 procedures procedure NOUN bracis-28409 135 15 of of ADP bracis-28409 135 16 this this DET bracis-28409 135 17 paper paper NOUN bracis-28409 135 18 are be AUX bracis-28409 135 19 implemented implement VERB bracis-28409 135 20 using use VERB bracis-28409 135 21 the the DET bracis-28409 135 22 scikit scikit NOUN bracis-28409 135 23 - - PUNCT bracis-28409 135 24 learn learn VERB bracis-28409 135 25 package package NOUN bracis-28409 135 26 available available ADJ bracis-28409 135 27 in in ADP bracis-28409 135 28 the the DET bracis-28409 135 29 python python NOUN bracis-28409 135 30 language language NOUN bracis-28409 135 31 . . PUNCT bracis-28409 136 1 in in ADP bracis-28409 136 2 order order NOUN bracis-28409 136 3 to to PART bracis-28409 136 4 create create VERB bracis-28409 136 5 the the DET bracis-28409 136 6 algorithm algorithm NOUN bracis-28409 136 7 proposed propose VERB bracis-28409 136 8 in in ADP bracis-28409 136 9 this this DET bracis-28409 136 10 article article NOUN bracis-28409 136 11 , , PUNCT bracis-28409 136 12 we we PRON bracis-28409 136 13 decided decide VERB bracis-28409 136 14 to to PART bracis-28409 136 15 use use VERB bracis-28409 136 16 the the DET bracis-28409 136 17 scikit scikit NOUN bracis-28409 136 18 - - PUNCT bracis-28409 136 19 learn learn VERB bracis-28409 136 20 python python NOUN bracis-28409 136 21 library library NOUN bracis-28409 137 1 [ [ X bracis-28409 137 2 23 23 NUM bracis-28409 137 3 ] ] PUNCT bracis-28409 137 4 , , PUNCT bracis-28409 137 5 since since SCONJ bracis-28409 137 6 it it PRON bracis-28409 137 7 offers offer VERB bracis-28409 137 8 several several ADJ bracis-28409 137 9 learning learning NOUN bracis-28409 137 10 methods method NOUN bracis-28409 137 11 for for ADP bracis-28409 137 12 classification classification NOUN bracis-28409 137 13 , , PUNCT bracis-28409 137 14 regression regression NOUN bracis-28409 137 15 , , PUNCT bracis-28409 137 16 and and CCONJ bracis-28409 137 17 clustering cluster VERB bracis-28409 137 18 domains domain NOUN bracis-28409 137 19 . . PUNCT bracis-28409 138 1 it it PRON bracis-28409 138 2 also also ADV bracis-28409 138 3 integrates integrate VERB bracis-28409 138 4 with with ADP bracis-28409 138 5 other other ADJ bracis-28409 138 6 widely widely ADV bracis-28409 138 7 used use VERB bracis-28409 138 8 python python NOUN bracis-28409 138 9 libraries library NOUN bracis-28409 138 10 for for ADP bracis-28409 138 11 machine machine NOUN bracis-28409 138 12 learning learn VERB bracis-28409 138 13 experiments experiment NOUN bracis-28409 138 14 , , PUNCT bracis-28409 138 15 such such ADJ bracis-28409 138 16 as as ADP bracis-28409 138 17 matplotlib matplotlib PROPN bracis-28409 138 18 [ [ X bracis-28409 138 19 24 24 NUM bracis-28409 138 20 ] ] PUNCT bracis-28409 138 21 , , PUNCT bracis-28409 138 22 numpy numpy NOUN bracis-28409 139 1 [ [ X bracis-28409 139 2 25 25 NUM bracis-28409 139 3 ] ] PUNCT bracis-28409 139 4 , , PUNCT bracis-28409 139 5 pandas panda VERB bracis-28409 139 6 [ [ X bracis-28409 139 7 26 26 NUM bracis-28409 139 8 ] ] PUNCT bracis-28409 139 9 , , PUNCT bracis-28409 139 10 and and CCONJ bracis-28409 139 11 others other NOUN bracis-28409 139 12 . . PUNCT bracis-28409 140 1 in in ADP bracis-28409 140 2 order order NOUN bracis-28409 140 3 to to PART bracis-28409 140 4 facilitate facilitate VERB bracis-28409 140 5 the the DET bracis-28409 140 6 comparative comparative ADJ bracis-28409 140 7 analysis analysis NOUN bracis-28409 140 8 of of ADP bracis-28409 140 9 the the DET bracis-28409 140 10 results result NOUN bracis-28409 140 11 of of ADP bracis-28409 140 12 the the DET bracis-28409 140 13 flexcon flexcon NOUN bracis-28409 140 14 - - PUNCT bracis-28409 140 15 ce ce PROPN bracis-28409 140 16 method method NOUN bracis-28409 140 17 with with ADP bracis-28409 140 18 flexcon flexcon NOUN bracis-28409 140 19 - - PUNCT bracis-28409 140 20 c c NOUN bracis-28409 140 21 , , PUNCT bracis-28409 140 22 three three NUM bracis-28409 140 23 classifiers classifier NOUN bracis-28409 140 24 are be AUX bracis-28409 140 25 used use VERB bracis-28409 140 26 , , PUNCT bracis-28409 140 27 which which PRON bracis-28409 140 28 are be AUX bracis-28409 140 29 : : PUNCT bracis-28409 140 30 naive naive ADJ bracis-28409 140 31 bayes bayes NOUN bracis-28409 140 32 ( ( PUNCT bracis-28409 140 33 nb nb NOUN bracis-28409 140 34 ) ) PUNCT bracis-28409 140 35 , , PUNCT bracis-28409 140 36 decision decision NOUN bracis-28409 140 37 tree tree NOUN bracis-28409 140 38 ( ( PUNCT bracis-28409 140 39 tree tree NOUN bracis-28409 140 40 ) ) PUNCT bracis-28409 140 41 – – PUNCT bracis-28409 140 42 using use VERB bracis-28409 140 43 a a DET bracis-28409 140 44 cart cart NOUN bracis-28409 140 45 - - PUNCT bracis-28409 140 46 based base VERB bracis-28409 140 47 algorithm algorithm NOUN bracis-28409 140 48 – – PUNCT bracis-28409 140 49 and and CCONJ bracis-28409 140 50 k k X bracis-28409 140 51 - - PUNCT bracis-28409 140 52 nearest near ADJ bracis-28409 140 53 neighbors neighbor NOUN bracis-28409 140 54 ( ( PUNCT bracis-28409 140 55 k k X bracis-28409 140 56 - - PUNCT bracis-28409 140 57 nn nn NOUN bracis-28409 140 58 ) ) PUNCT bracis-28409 140 59 – – PUNCT bracis-28409 140 60 with with ADP bracis-28409 140 61 k k PROPN bracis-28409 140 62 values value NOUN bracis-28409 140 63 from from ADP bracis-28409 140 64 4 4 NUM bracis-28409 140 65 to to PART bracis-28409 140 66 8 8 NUM bracis-28409 140 67 . . PUNCT bracis-28409 141 1 additionally additionally ADV bracis-28409 141 2 , , PUNCT bracis-28409 141 3 the the DET bracis-28409 141 4 percentage percentage NOUN bracis-28409 141 5 values value NOUN bracis-28409 141 6 of of ADP bracis-28409 141 7 the the DET bracis-28409 141 8 initially initially ADV bracis-28409 141 9 labeled label VERB bracis-28409 141 10 instances instance NOUN bracis-28409 141 11 , , PUNCT bracis-28409 141 12 n n CCONJ bracis-28409 141 13 , , PUNCT bracis-28409 141 14 are be AUX bracis-28409 141 15 5\(\%\ 5\(\%\ NUM bracis-28409 141 16 ) ) PUNCT bracis-28409 141 17 , , PUNCT bracis-28409 141 18 10\(\%\ 10\(\%\ NUM bracis-28409 141 19 ) ) PUNCT bracis-28409 141 20 , , PUNCT bracis-28409 141 21 15\(\%\ 15\(\%\ NUM bracis-28409 141 22 ) ) PUNCT bracis-28409 141 23 , , PUNCT bracis-28409 141 24 20\(\%\ 20\(\%\ NUM bracis-28409 141 25 ) ) PUNCT bracis-28409 141 26 and and CCONJ bracis-28409 141 27 25\(\%\ 25\(\%\ NOUN bracis-28409 141 28 ) ) PUNCT bracis-28409 141 29 , , PUNCT bracis-28409 141 30 in in ADP bracis-28409 141 31 the the DET bracis-28409 141 32 same same ADJ bracis-28409 141 33 way way NOUN bracis-28409 141 34 as as SCONJ bracis-28409 141 35 it it PRON bracis-28409 141 36 was be AUX bracis-28409 141 37 performed perform VERB bracis-28409 141 38 by by ADP bracis-28409 141 39 [ [ X bracis-28409 141 40 16 16 NUM bracis-28409 141 41 , , PUNCT bracis-28409 141 42 19 19 NUM bracis-28409 141 43 , , PUNCT bracis-28409 141 44 27 27 NUM bracis-28409 141 45 , , PUNCT bracis-28409 141 46 28 28 NUM bracis-28409 141 47 ] ] PUNCT bracis-28409 141 48 . . PUNCT bracis-28409 142 1 each each DET bracis-28409 142 2 trained train VERB bracis-28409 142 3 classifier classifier NOUN bracis-28409 142 4 was be AUX bracis-28409 142 5 included include VERB bracis-28409 142 6 to to ADP bracis-28409 142 7 the the DET bracis-28409 142 8 classifier classifier NOUN bracis-28409 142 9 ensemble ensemble VERB bracis-28409 142 10 heterogeneously heterogeneously ADV bracis-28409 142 11 and and CCONJ bracis-28409 142 12 homogeneously homogeneously ADV bracis-28409 142 13 , , PUNCT bracis-28409 142 14 for for ADP bracis-28409 142 15 flexcon flexcon NOUN bracis-28409 142 16 - - PUNCT bracis-28409 142 17 ce ce PROPN bracis-28409 142 18 ( ( PUNCT bracis-28409 142 19 het het PROPN bracis-28409 142 20 ) ) PUNCT bracis-28409 142 21 and and CCONJ bracis-28409 142 22 flexcon flexcon NOUN bracis-28409 142 23 - - PUNCT bracis-28409 142 24 ce ce PROPN bracis-28409 142 25 ( ( PUNCT bracis-28409 142 26 hom hom NOUN bracis-28409 142 27 ) ) PUNCT bracis-28409 142 28 , , PUNCT bracis-28409 142 29 respectively respectively ADV bracis-28409 142 30 . . PUNCT bracis-28409 143 1 for for ADP bracis-28409 143 2 the the DET bracis-28409 143 3 combination combination NOUN bracis-28409 143 4 of of ADP bracis-28409 143 5 all all PRON bracis-28409 143 6 n n DET bracis-28409 143 7 classifiers classifier NOUN bracis-28409 143 8 , , PUNCT bracis-28409 143 9 the the DET bracis-28409 143 10 voting voting NOUN bracis-28409 143 11 criterion criterion NOUN bracis-28409 143 12 was be AUX bracis-28409 143 13 applied apply VERB bracis-28409 143 14 to to PART bracis-28409 143 15 define define VERB bracis-28409 143 16 the the DET bracis-28409 143 17 overall overall ADJ bracis-28409 143 18 ensemble ensemble ADJ bracis-28409 143 19 output output NOUN bracis-28409 143 20 . . PUNCT bracis-28409 144 1 the the DET bracis-28409 144 2 voting voting NOUN bracis-28409 144 3 method method NOUN bracis-28409 144 4 consists consist VERB bracis-28409 144 5 of of ADP bracis-28409 144 6 a a DET bracis-28409 144 7 non non ADJ bracis-28409 144 8 - - ADJ bracis-28409 144 9 linear linear ADJ bracis-28409 144 10 combination combination NOUN bracis-28409 144 11 of of ADP bracis-28409 144 12 the the DET bracis-28409 144 13 classifiers classifier NOUN bracis-28409 144 14 ’ ’ PART bracis-28409 144 15 outputs output NOUN bracis-28409 144 16 , , PUNCT bracis-28409 144 17 and and CCONJ bracis-28409 144 18 its its PRON bracis-28409 144 19 process process NOUN bracis-28409 144 20 consists consist VERB bracis-28409 144 21 of of ADP bracis-28409 144 22 determining determine VERB bracis-28409 144 23 the the DET bracis-28409 144 24 winning win VERB bracis-28409 144 25 class class NOUN bracis-28409 144 26 , , PUNCT bracis-28409 144 27 from from ADP bracis-28409 144 28 a a DET bracis-28409 144 29 data data NOUN bracis-28409 144 30 input input NOUN bracis-28409 144 31 pattern pattern NOUN bracis-28409 144 32 , , PUNCT bracis-28409 144 33 by by ADP bracis-28409 144 34 the the DET bracis-28409 144 35 total total ADJ bracis-28409 144 36 number number NOUN bracis-28409 144 37 of of ADP bracis-28409 144 38 votes vote NOUN bracis-28409 144 39 counted count VERB bracis-28409 144 40 for for ADP bracis-28409 144 41 each each DET bracis-28409 144 42 classifier classifier NOUN bracis-28409 144 43 [ [ X bracis-28409 144 44 4 4 NUM bracis-28409 144 45 , , PUNCT bracis-28409 144 46 8 8 NUM bracis-28409 144 47 , , PUNCT bracis-28409 144 48 29 29 NUM bracis-28409 144 49 ] ] PUNCT bracis-28409 144 50 . . PUNCT bracis-28409 145 1 therefore therefore ADV bracis-28409 145 2 , , PUNCT bracis-28409 145 3 the the DET bracis-28409 145 4 confidence confidence NOUN bracis-28409 145 5 of of ADP bracis-28409 145 6 the the DET bracis-28409 145 7 classifier classifier NOUN bracis-28409 145 8 ensemble ensemble ADJ bracis-28409 145 9 is be AUX bracis-28409 145 10 defined define VERB bracis-28409 145 11 by by ADP bracis-28409 145 12 the the DET bracis-28409 145 13 number number NOUN bracis-28409 145 14 of of ADP bracis-28409 145 15 classifiers classifier NOUN bracis-28409 145 16 that that PRON bracis-28409 145 17 selected select VERB bracis-28409 145 18 the the DET bracis-28409 145 19 winning win VERB bracis-28409 145 20 class class NOUN bracis-28409 145 21 divided divide VERB bracis-28409 145 22 by by ADP bracis-28409 145 23 the the DET bracis-28409 145 24 number number NOUN bracis-28409 145 25 of of ADP bracis-28409 145 26 classifiers classifier NOUN bracis-28409 145 27 of of ADP bracis-28409 145 28 the the DET bracis-28409 145 29 used use VERB bracis-28409 145 30 ensemble ensemble NOUN bracis-28409 145 31 . . PUNCT bracis-28409 146 1 the the DET bracis-28409 146 2 main main ADJ bracis-28409 146 3 aim aim NOUN bracis-28409 146 4 of of ADP bracis-28409 146 5 this this DET bracis-28409 146 6 empirical empirical ADJ bracis-28409 146 7 analysis analysis NOUN bracis-28409 146 8 is be AUX bracis-28409 146 9 to to PART bracis-28409 146 10 evaluate evaluate VERB bracis-28409 146 11 the the DET bracis-28409 146 12 performance performance NOUN bracis-28409 146 13 of of ADP bracis-28409 146 14 the the DET bracis-28409 146 15 flexcon flexcon NOUN bracis-28409 146 16 - - PUNCT bracis-28409 146 17 ce ce NOUN bracis-28409 146 18 algorithm algorithm NOUN bracis-28409 146 19 and and CCONJ bracis-28409 146 20 to to PART bracis-28409 146 21 compare compare VERB bracis-28409 146 22 them they PRON bracis-28409 146 23 to to ADP bracis-28409 146 24 the the DET bracis-28409 146 25 flexcon flexcon NOUN bracis-28409 146 26 - - PUNCT bracis-28409 146 27 c c NOUN bracis-28409 146 28 ( ( PUNCT bracis-28409 146 29 previously previously ADV bracis-28409 146 30 explained explain VERB bracis-28409 146 31 : : PUNCT bracis-28409 146 32 flexcon flexcon NOUN bracis-28409 146 33 - - PUNCT bracis-28409 146 34 c1(v c1(v NOUN bracis-28409 146 35 ) ) PUNCT bracis-28409 146 36 , , PUNCT bracis-28409 146 37 flexcon flexcon NOUN bracis-28409 146 38 - - PUNCT bracis-28409 146 39 c1(s c1(s NOUN bracis-28409 146 40 ) ) PUNCT bracis-28409 146 41 and and CCONJ bracis-28409 146 42 flexcon flexcon NOUN bracis-28409 146 43 - - PUNCT bracis-28409 146 44 c2 c2 PROPN bracis-28409 146 45 ) ) PUNCT bracis-28409 146 46 results result NOUN bracis-28409 146 47 . . PUNCT bracis-28409 147 1 in in ADP bracis-28409 147 2 order order NOUN bracis-28409 147 3 to to PART bracis-28409 147 4 assess assess VERB bracis-28409 147 5 the the DET bracis-28409 147 6 obtained obtain VERB bracis-28409 147 7 results result NOUN bracis-28409 147 8 , , PUNCT bracis-28409 147 9 the the DET bracis-28409 147 10 friedman friedman PROPN bracis-28409 147 11 statistical statistical ADJ bracis-28409 147 12 test test NOUN bracis-28409 147 13 was be AUX bracis-28409 147 14 achieved achieve VERB bracis-28409 147 15 . . PUNCT bracis-28409 148 1 according accord VERB bracis-28409 148 2 to to ADP bracis-28409 148 3 [ [ X bracis-28409 148 4 30 30 NUM bracis-28409 148 5 ] ] PUNCT bracis-28409 148 6 , , PUNCT bracis-28409 148 7 this this DET bracis-28409 148 8 test test NOUN bracis-28409 148 9 compares compare VERB bracis-28409 148 10 related related ADJ bracis-28409 148 11 sample sample NOUN bracis-28409 148 12 data datum NOUN bracis-28409 148 13 , , PUNCT bracis-28409 148 14 which which PRON bracis-28409 148 15 means mean VERB bracis-28409 148 16 that that SCONJ bracis-28409 148 17 the the DET bracis-28409 148 18 same same ADJ bracis-28409 148 19 object object NOUN bracis-28409 148 20 can can AUX bracis-28409 148 21 be be AUX bracis-28409 148 22 analyzed analyze VERB bracis-28409 148 23 more more ADJ bracis-28409 148 24 than than ADP bracis-28409 148 25 once once ADV bracis-28409 148 26 . . PUNCT bracis-28409 149 1 as as SCONJ bracis-28409 149 2 it it PRON bracis-28409 149 3 is be AUX bracis-28409 149 4 non non ADJ bracis-28409 149 5 - - ADJ bracis-28409 149 6 parametric parametric ADJ bracis-28409 149 7 test test NOUN bracis-28409 149 8 , , PUNCT bracis-28409 149 9 numerical numerical ADJ bracis-28409 149 10 values value NOUN bracis-28409 149 11 are be AUX bracis-28409 149 12 not not PART bracis-28409 149 13 used use VERB bracis-28409 149 14 directly directly ADV bracis-28409 149 15 , , PUNCT bracis-28409 149 16 but but CCONJ bracis-28409 149 17 rather rather ADV bracis-28409 149 18 their their PRON bracis-28409 149 19 positions position NOUN bracis-28409 149 20 . . PUNCT bracis-28409 150 1 after after SCONJ bracis-28409 150 2 the the DET bracis-28409 150 3 ranking ranking NOUN bracis-28409 150 4 are be AUX bracis-28409 150 5 performed perform VERB bracis-28409 150 6 for for ADP bracis-28409 150 7 the the DET bracis-28409 150 8 groups group NOUN bracis-28409 150 9 separately separately ADV bracis-28409 150 10 , , PUNCT bracis-28409 150 11 the the DET bracis-28409 150 12 hypothesis hypothesis NOUN bracis-28409 150 13 of of ADP bracis-28409 150 14 similarity similarity NOUN bracis-28409 150 15 of of ADP bracis-28409 150 16 the the DET bracis-28409 150 17 sum sum NOUN bracis-28409 150 18 of of ADP bracis-28409 150 19 the the DET bracis-28409 150 20 position position NOUN bracis-28409 150 21 of of ADP bracis-28409 150 22 each each DET bracis-28409 150 23 group group NOUN bracis-28409 150 24 is be AUX bracis-28409 150 25 verified verify VERB bracis-28409 150 26 . . PUNCT bracis-28409 151 1 in in ADP bracis-28409 151 2 this this DET bracis-28409 151 3 way way NOUN bracis-28409 151 4 , , PUNCT bracis-28409 151 5 it it PRON bracis-28409 151 6 is be AUX bracis-28409 151 7 possible possible ADJ bracis-28409 151 8 to to PART bracis-28409 151 9 compare compare VERB bracis-28409 151 10 the the DET bracis-28409 151 11 performance performance NOUN bracis-28409 151 12 of of ADP bracis-28409 151 13 different different ADJ bracis-28409 151 14 data datum NOUN bracis-28409 151 15 classification classification NOUN bracis-28409 151 16 algorithms algorithm NOUN bracis-28409 151 17 in in ADP bracis-28409 151 18 the the DET bracis-28409 151 19 scenarios scenario NOUN bracis-28409 151 20 created create VERB bracis-28409 151 21 for for ADP bracis-28409 151 22 each each DET bracis-28409 151 23 different different ADJ bracis-28409 151 24 dataset dataset NOUN bracis-28409 151 25 . . PUNCT bracis-28409 152 1 6 6 NUM bracis-28409 152 2 experimental experimental ADJ bracis-28409 152 3 results result NOUN bracis-28409 152 4 this this DET bracis-28409 152 5 section section NOUN bracis-28409 152 6 presents present VERB bracis-28409 152 7 the the DET bracis-28409 152 8 analysis analysis NOUN bracis-28409 152 9 of of ADP bracis-28409 152 10 the the DET bracis-28409 152 11 obtained obtain VERB bracis-28409 152 12 results result NOUN bracis-28409 152 13 . . PUNCT bracis-28409 153 1 as as SCONJ bracis-28409 153 2 previously previously ADV bracis-28409 153 3 mentioned mention VERB bracis-28409 153 4 , , PUNCT bracis-28409 153 5 the the DET bracis-28409 153 6 analysis analysis NOUN bracis-28409 153 7 of of ADP bracis-28409 153 8 the the DET bracis-28409 153 9 results result NOUN bracis-28409 153 10 will will AUX bracis-28409 153 11 be be AUX bracis-28409 153 12 conducted conduct VERB bracis-28409 153 13 in in ADP bracis-28409 153 14 a a DET bracis-28409 153 15 comparative comparative ADJ bracis-28409 153 16 way way NOUN bracis-28409 153 17 , , PUNCT bracis-28409 153 18 evaluating evaluate VERB bracis-28409 153 19 the the DET bracis-28409 153 20 performance performance NOUN bracis-28409 153 21 of of ADP bracis-28409 153 22 both both DET bracis-28409 153 23 flexcon flexcon NOUN bracis-28409 153 24 - - PUNCT bracis-28409 153 25 ce ce NOUN bracis-28409 153 26 versions version NOUN bracis-28409 153 27 and and CCONJ bracis-28409 153 28 three three NUM bracis-28409 153 29 flexcon flexcon NOUN bracis-28409 153 30 - - PUNCT bracis-28409 153 31 c c NOUN bracis-28409 153 32 versions version NOUN bracis-28409 153 33 , , PUNCT bracis-28409 153 34 flexcon flexcon NOUN bracis-28409 153 35 - - PUNCT bracis-28409 153 36 c1(v c1(v NOUN bracis-28409 153 37 ) ) PUNCT bracis-28409 153 38 , , PUNCT bracis-28409 153 39 flexcon flexcon NOUN bracis-28409 153 40 - - PUNCT bracis-28409 153 41 c1(s c1(s NOUN bracis-28409 153 42 ) ) PUNCT bracis-28409 153 43 , , PUNCT bracis-28409 153 44 and and CCONJ bracis-28409 153 45 flexcon flexcon NOUN bracis-28409 153 46 - - PUNCT bracis-28409 153 47 c2 c2 PROPN bracis-28409 153 48 [ [ X bracis-28409 153 49 3 3 NUM bracis-28409 153 50 ] ] PUNCT bracis-28409 153 51 . . PUNCT bracis-28409 154 1 the the DET bracis-28409 154 2 next next ADJ bracis-28409 154 3 sections section NOUN bracis-28409 154 4 will will AUX bracis-28409 154 5 present present VERB bracis-28409 154 6 the the DET bracis-28409 154 7 analysis analysis NOUN bracis-28409 154 8 of of ADP bracis-28409 154 9 the the DET bracis-28409 154 10 experimental experimental ADJ bracis-28409 154 11 results result NOUN bracis-28409 154 12 . . PUNCT bracis-28409 155 1 the the DET bracis-28409 155 2 first first ADJ bracis-28409 155 3 one one NUM bracis-28409 155 4 presents present VERB bracis-28409 155 5 an an DET bracis-28409 155 6 analysis analysis NOUN bracis-28409 155 7 of of ADP bracis-28409 155 8 the the DET bracis-28409 155 9 accuracy accuracy NOUN bracis-28409 155 10 results result NOUN bracis-28409 155 11 presented present VERB bracis-28409 155 12 by by ADP bracis-28409 155 13 all all PRON bracis-28409 155 14 analyzed analyze VERB bracis-28409 155 15 methods.while methods.while ADJ bracis-28409 155 16 the the DET bracis-28409 155 17 next next ADJ bracis-28409 155 18 section section NOUN bracis-28409 155 19 conducts conduct VERB bracis-28409 155 20 a a DET bracis-28409 155 21 statistical statistical ADJ bracis-28409 155 22 analysis analysis NOUN bracis-28409 155 23 of of ADP bracis-28409 155 24 the the DET bracis-28409 155 25 obtained obtain VERB bracis-28409 155 26 results result NOUN bracis-28409 155 27 . . PUNCT bracis-28409 156 1 6.1 6.1 NUM bracis-28409 156 2 accuracy accuracy NOUN bracis-28409 156 3 analysis analysis NOUN bracis-28409 156 4 tables table NOUN bracis-28409 156 5     SPACE bracis-28409 156 6 2 2 NUM bracis-28409 156 7 and and CCONJ bracis-28409 156 8 3 3 NUM bracis-28409 156 9 present present NOUN bracis-28409 156 10 the the DET bracis-28409 156 11 results result NOUN bracis-28409 156 12 obtained obtain VERB bracis-28409 156 13 by by ADP bracis-28409 156 14 each each DET bracis-28409 156 15 method method NOUN bracis-28409 156 16 used use VERB bracis-28409 156 17 in in ADP bracis-28409 156 18 this this DET bracis-28409 156 19 analysis analysis NOUN bracis-28409 156 20 . . PUNCT bracis-28409 157 1 the the DET bracis-28409 157 2 difference difference NOUN bracis-28409 157 3 between between ADP bracis-28409 157 4 these these DET bracis-28409 157 5 two two NUM bracis-28409 157 6 tables table NOUN bracis-28409 157 7 is be AUX bracis-28409 157 8 that that SCONJ bracis-28409 157 9 the the DET bracis-28409 157 10 first first ADJ bracis-28409 157 11 table table NOUN bracis-28409 157 12 shows show VERB bracis-28409 157 13 the the DET bracis-28409 157 14 results result NOUN bracis-28409 157 15 of of ADP bracis-28409 157 16 three three NUM bracis-28409 157 17 different different ADJ bracis-28409 157 18 classifiers classifier NOUN bracis-28409 157 19 and and CCONJ bracis-28409 157 20 compares compare VERB bracis-28409 157 21 the the DET bracis-28409 157 22 performance performance NOUN bracis-28409 157 23 of of ADP bracis-28409 157 24 flexcon flexcon NOUN bracis-28409 157 25 - - PUNCT bracis-28409 157 26 ce ce NOUN bracis-28409 157 27 using use VERB bracis-28409 157 28 a a DET bracis-28409 157 29 homogeneous homogeneous ADJ bracis-28409 157 30 classifier classifier NOUN bracis-28409 157 31 ensemble ensemble ADJ bracis-28409 157 32 ( ( PUNCT bracis-28409 157 33 flexcon flexcon NOUN bracis-28409 157 34 - - PUNCT bracis-28409 157 35 ce(hom ce(hom NOUN bracis-28409 157 36 ) ) PUNCT bracis-28409 157 37 ) ) PUNCT bracis-28409 157 38 with with ADP bracis-28409 157 39 all all DET bracis-28409 157 40 three three NUM bracis-28409 157 41 flexcon flexcon ADJ bracis-28409 157 42 - - PUNCT bracis-28409 157 43 c c NOUN bracis-28409 157 44 versions version NOUN bracis-28409 157 45 ( ( PUNCT bracis-28409 157 46 flexcon flexcon NOUN bracis-28409 157 47 - - PUNCT bracis-28409 157 48 c1(v c1(v NOUN bracis-28409 157 49 ) ) PUNCT bracis-28409 157 50 , , PUNCT bracis-28409 157 51 flexcon flexcon NOUN bracis-28409 157 52 - - PUNCT bracis-28409 157 53 c1(s c1(s NOUN bracis-28409 157 54 ) ) PUNCT bracis-28409 157 55 , , PUNCT bracis-28409 157 56 and and CCONJ bracis-28409 157 57 flexcon flexcon NOUN bracis-28409 157 58 - - PUNCT bracis-28409 157 59 c2 c2 PROPN bracis-28409 157 60 ) ) PUNCT bracis-28409 157 61 . . PUNCT bracis-28409 158 1 for for ADP bracis-28409 158 2 the the DET bracis-28409 158 3 second second ADJ bracis-28409 158 4 table table NOUN bracis-28409 158 5 , , PUNCT bracis-28409 158 6 the the DET bracis-28409 158 7 best good ADJ bracis-28409 158 8 results result NOUN bracis-28409 158 9 of of ADP bracis-28409 158 10 table table NOUN bracis-28409 158 11     SPACE bracis-28409 158 12 2 2 NUM bracis-28409 158 13 are be AUX bracis-28409 158 14 selected select VERB bracis-28409 158 15 , , PUNCT bracis-28409 158 16 for for ADP bracis-28409 158 17 each each DET bracis-28409 158 18 classification classification NOUN bracis-28409 158 19 algorithm algorithm NOUN bracis-28409 158 20 ( ( PUNCT bracis-28409 158 21 naive naive ADJ bracis-28409 158 22 bayes bayes NOUN bracis-28409 158 23 , , PUNCT bracis-28409 158 24 decision decision NOUN bracis-28409 158 25 tree tree NOUN bracis-28409 158 26 and and CCONJ bracis-28409 158 27 k k PROPN bracis-28409 158 28 - - PUNCT bracis-28409 158 29 nn nn NOUN bracis-28409 158 30 ) ) PUNCT bracis-28409 158 31 , , PUNCT bracis-28409 158 32 as as ADV bracis-28409 158 33 well well ADV bracis-28409 158 34 as as ADP bracis-28409 158 35 the the DET bracis-28409 158 36 flexcon flexcon NOUN bracis-28409 158 37 - - PUNCT bracis-28409 158 38 ce ce PROPN bracis-28409 158 39 result result NOUN bracis-28409 158 40 using use VERB bracis-28409 158 41 a a DET bracis-28409 158 42 heterogeneous heterogeneous ADJ bracis-28409 158 43 classifier classifier NOUN bracis-28409 158 44 ensemble ensemble ADJ bracis-28409 158 45 ( ( PUNCT bracis-28409 158 46 flexcon flexcon NOUN bracis-28409 158 47 - - PUNCT bracis-28409 158 48 ce(het ce(het NOUN bracis-28409 158 49 ) ) PUNCT bracis-28409 158 50 ) ) PUNCT bracis-28409 158 51 . . PUNCT bracis-28409 159 1 table table NOUN bracis-28409 160 1 2 2 NUM bracis-28409 160 2 . . PUNCT bracis-28409 160 3 results result NOUN bracis-28409 160 4 of of ADP bracis-28409 160 5 average average ADJ bracis-28409 160 6 accuraryfull accuraryfull ADJ bracis-28409 160 7 size size NOUN bracis-28409 160 8 table table NOUN bracis-28409 160 9 table table NOUN bracis-28409 160 10 3 3 NUM bracis-28409 160 11 . . PUNCT bracis-28409 160 12 results result NOUN bracis-28409 160 13 of of ADP bracis-28409 160 14 best good ADJ bracis-28409 160 15 average average ADJ bracis-28409 160 16 accuracyfull accuracyfull ADJ bracis-28409 160 17 size size NOUN bracis-28409 160 18 table table NOUN bracis-28409 160 19 both both DET bracis-28409 160 20 tables table NOUN bracis-28409 160 21 have have VERB bracis-28409 160 22 the the DET bracis-28409 160 23 same same ADJ bracis-28409 160 24 configuration configuration NOUN bracis-28409 160 25 : : PUNCT bracis-28409 160 26 the the DET bracis-28409 160 27 first first ADJ bracis-28409 160 28 column column NOUN bracis-28409 160 29 presents present VERB bracis-28409 160 30 the the DET bracis-28409 160 31 name name NOUN bracis-28409 160 32 of of ADP bracis-28409 160 33 the the DET bracis-28409 160 34 used use VERB bracis-28409 160 35 classifier classifier NOUN bracis-28409 160 36 while while SCONJ bracis-28409 160 37 the the DET bracis-28409 160 38 second second ADJ bracis-28409 160 39 column column NOUN bracis-28409 160 40 indicates indicate VERB bracis-28409 160 41 the the DET bracis-28409 160 42 semi semi ADJ bracis-28409 160 43 - - ADJ bracis-28409 160 44 supervised supervised ADJ bracis-28409 160 45 learning learning NOUN bracis-28409 160 46 method method NOUN bracis-28409 160 47 . . PUNCT bracis-28409 161 1 columns column NOUN bracis-28409 161 2 3–7 3–7 NUM bracis-28409 161 3 display display VERB bracis-28409 161 4 the the DET bracis-28409 161 5 average average ADJ bracis-28409 161 6 accuracy accuracy NOUN bracis-28409 161 7 and and CCONJ bracis-28409 161 8 standard standard ADJ bracis-28409 161 9 deviation deviation NOUN bracis-28409 161 10 obtained obtain VERB bracis-28409 161 11 by by ADP bracis-28409 161 12 each each DET bracis-28409 161 13 semi semi ADJ bracis-28409 161 14 - - ADJ bracis-28409 161 15 supervised supervised ADJ bracis-28409 161 16 method method NOUN bracis-28409 161 17 over over ADP bracis-28409 161 18 all all DET bracis-28409 161 19 20 20 NUM bracis-28409 161 20 used use VERB bracis-28409 161 21 datasets dataset NOUN bracis-28409 161 22 , , PUNCT bracis-28409 161 23 according accord VERB bracis-28409 161 24 to to ADP bracis-28409 161 25 the the DET bracis-28409 161 26 settings setting NOUN bracis-28409 161 27 of of ADP bracis-28409 161 28 5\(\%\ 5\(\%\ NUM bracis-28409 161 29 ) ) PUNCT bracis-28409 161 30 , , PUNCT bracis-28409 161 31 10\(\%\ 10\(\%\ NUM bracis-28409 161 32 ) ) PUNCT bracis-28409 161 33 , , PUNCT bracis-28409 161 34 15\(\%\ 15\(\%\ NUM bracis-28409 161 35 ) ) PUNCT bracis-28409 161 36 , , PUNCT bracis-28409 161 37 20\(\%\ 20\(\%\ NUM bracis-28409 161 38 ) ) PUNCT bracis-28409 161 39 , , PUNCT bracis-28409 161 40 and and CCONJ bracis-28409 161 41 25\(\%\ 25\(\%\ NUM bracis-28409 161 42 ) ) PUNCT bracis-28409 161 43 of of ADP bracis-28409 161 44 initially initially ADV bracis-28409 161 45 labeled label VERB bracis-28409 161 46 instances instance NOUN bracis-28409 161 47 . . PUNCT bracis-28409 162 1 additionally additionally ADV bracis-28409 162 2 , , PUNCT bracis-28409 162 3 in in ADP bracis-28409 162 4 both both DET bracis-28409 162 5 tables table NOUN bracis-28409 162 6 , , PUNCT bracis-28409 162 7 the the DET bracis-28409 162 8 results result NOUN bracis-28409 162 9 that that PRON bracis-28409 162 10 showed show VERB bracis-28409 162 11 the the DET bracis-28409 162 12 most most ADV bracis-28409 162 13 significant significant ADJ bracis-28409 162 14 accuracy accuracy NOUN bracis-28409 162 15 values value NOUN bracis-28409 162 16 are be AUX bracis-28409 162 17 highlighted highlight VERB bracis-28409 162 18 in in ADP bracis-28409 162 19 bold bold ADJ bracis-28409 162 20 . . PUNCT bracis-28409 163 1 when when SCONJ bracis-28409 163 2 analyzing analyzing NOUN bracis-28409 163 3 results result NOUN bracis-28409 163 4 obtained obtain VERB bracis-28409 163 5 with with ADP bracis-28409 163 6 the the DET bracis-28409 163 7 naive naive ADJ bracis-28409 163 8 bayes bayes NOUN bracis-28409 163 9 classifier classifier NOUN bracis-28409 163 10 in in ADP bracis-28409 163 11 table table NOUN bracis-28409 163 12     SPACE bracis-28409 163 13 2 2 NUM bracis-28409 163 14 , , PUNCT bracis-28409 163 15 it it PRON bracis-28409 163 16 is be AUX bracis-28409 163 17 possible possible ADJ bracis-28409 163 18 to to PART bracis-28409 163 19 observe observe VERB bracis-28409 163 20 that that SCONJ bracis-28409 163 21 the the DET bracis-28409 163 22 flexcon flexcon NOUN bracis-28409 163 23 - - PUNCT bracis-28409 163 24 ce ce PROPN bracis-28409 163 25 method method NOUN bracis-28409 163 26 using use VERB bracis-28409 163 27 the the DET bracis-28409 163 28 homogeneous homogeneous ADJ bracis-28409 163 29 ensemble ensemble ADJ bracis-28409 163 30 flexcon flexcon NOUN bracis-28409 163 31 - - PUNCT bracis-28409 163 32 ce(hom ce(hom NOUN bracis-28409 163 33 ) ) PUNCT bracis-28409 163 34 obtained obtain VERB bracis-28409 163 35 the the DET bracis-28409 163 36 best good ADJ bracis-28409 163 37 results result NOUN bracis-28409 163 38 in in ADP bracis-28409 163 39 80 80 NUM bracis-28409 163 40 % % NOUN bracis-28409 163 41 of of ADP bracis-28409 163 42 the the DET bracis-28409 163 43 analysed analyse VERB bracis-28409 163 44 cases case NOUN bracis-28409 163 45 ( ( PUNCT bracis-28409 163 46 4 4 NUM bracis-28409 163 47 out out ADP bracis-28409 163 48 of of ADP bracis-28409 163 49 the the DET bracis-28409 163 50 5 5 NUM bracis-28409 163 51 % % NOUN bracis-28409 163 52 of of ADP bracis-28409 163 53 initially initially ADV bracis-28409 163 54 labeled label VERB bracis-28409 163 55 instances instance NOUN bracis-28409 163 56 ) ) PUNCT bracis-28409 163 57 . . PUNCT bracis-28409 164 1 in in ADP bracis-28409 164 2 this this DET bracis-28409 164 3 same same ADJ bracis-28409 164 4 table table NOUN bracis-28409 164 5 , , PUNCT bracis-28409 164 6 considering consider VERB bracis-28409 164 7 the the DET bracis-28409 164 8 results result NOUN bracis-28409 164 9 obtained obtain VERB bracis-28409 164 10 by by ADP bracis-28409 164 11 the the DET bracis-28409 164 12 tree tree NOUN bracis-28409 164 13 classifier classifier NOUN bracis-28409 164 14 , , PUNCT bracis-28409 164 15 flexcon flexcon NOUN bracis-28409 164 16 - - PUNCT bracis-28409 164 17 ce(hom ce(hom NOUN bracis-28409 164 18 ) ) PUNCT bracis-28409 164 19 also also ADV bracis-28409 164 20 achieved achieve VERB bracis-28409 164 21 the the DET bracis-28409 164 22 highest high ADJ bracis-28409 164 23 accuracy accuracy NOUN bracis-28409 164 24 in in ADP bracis-28409 164 25 100 100 NUM bracis-28409 164 26 % % NOUN bracis-28409 164 27 of of ADP bracis-28409 164 28 cases case NOUN bracis-28409 164 29 ( ( PUNCT bracis-28409 164 30 5 5 NUM bracis-28409 164 31 out out ADP bracis-28409 164 32 of of ADP bracis-28409 164 33 the the DET bracis-28409 164 34 5 5 NUM bracis-28409 164 35 % % NOUN bracis-28409 164 36 of of ADP bracis-28409 164 37 initially initially ADV bracis-28409 164 38 labeled label VERB bracis-28409 164 39 data datum NOUN bracis-28409 164 40 ) ) PUNCT bracis-28409 164 41 . . PUNCT bracis-28409 165 1 based base VERB bracis-28409 165 2 on on ADP bracis-28409 165 3 these these DET bracis-28409 165 4 results result NOUN bracis-28409 165 5 , , PUNCT bracis-28409 165 6 it it PRON bracis-28409 165 7 is be AUX bracis-28409 165 8 possible possible ADJ bracis-28409 165 9 to to AUX bracis-28409 165 10 state state VERB bracis-28409 165 11 that that SCONJ bracis-28409 165 12 the the DET bracis-28409 165 13 flexcon flexcon NOUN bracis-28409 165 14 - - PUNCT bracis-28409 165 15 ce(hom ce(hom ADJ bracis-28409 165 16 ) ) PUNCT bracis-28409 165 17 method method NOUN bracis-28409 165 18 obtained obtain VERB bracis-28409 165 19 significantly significantly ADV bracis-28409 165 20 better well ADJ bracis-28409 165 21 performance performance NOUN bracis-28409 165 22 results result NOUN bracis-28409 165 23 in in ADP bracis-28409 165 24 two two NUM bracis-28409 165 25 out out ADP bracis-28409 165 26 of of ADP bracis-28409 165 27 three three NUM bracis-28409 165 28 analyzed analyze VERB bracis-28409 165 29 classifiers classifier NOUN bracis-28409 165 30 . . PUNCT bracis-28409 166 1 still still ADV bracis-28409 166 2 exploring explore VERB bracis-28409 166 3 table table NOUN bracis-28409 166 4     SPACE bracis-28409 166 5 2 2 NUM bracis-28409 166 6 , , PUNCT bracis-28409 166 7 in in ADP bracis-28409 166 8 general general ADJ bracis-28409 166 9 , , PUNCT bracis-28409 166 10 it it PRON bracis-28409 166 11 is be AUX bracis-28409 166 12 possible possible ADJ bracis-28409 166 13 to to PART bracis-28409 166 14 observe observe VERB bracis-28409 166 15 that that SCONJ bracis-28409 166 16 the the DET bracis-28409 166 17 flexcon flexcon NOUN bracis-28409 166 18 - - PUNCT bracis-28409 166 19 ce(hom ce(hom ADJ bracis-28409 166 20 ) ) PUNCT bracis-28409 166 21 method method NOUN bracis-28409 166 22 performed perform VERB bracis-28409 166 23 better well ADV bracis-28409 166 24 than than ADP bracis-28409 166 25 all all DET bracis-28409 166 26 flexcon flexcon NOUN bracis-28409 166 27 - - PUNCT bracis-28409 166 28 c c NOUN bracis-28409 166 29 versions version NOUN bracis-28409 166 30 in in ADP bracis-28409 166 31 60 60 NUM bracis-28409 166 32 % % NOUN bracis-28409 166 33 of of ADP bracis-28409 166 34 the the DET bracis-28409 166 35 analyzed analyze VERB bracis-28409 166 36 cases case NOUN bracis-28409 166 37 ( ( PUNCT bracis-28409 166 38 9 9 NUM bracis-28409 166 39 out out ADP bracis-28409 166 40 of of ADP bracis-28409 166 41 the the DET bracis-28409 166 42 15 15 NUM bracis-28409 166 43 best good ADJ bracis-28409 166 44 results result NOUN bracis-28409 166 45 ) ) PUNCT bracis-28409 166 46 . . PUNCT bracis-28409 167 1 this this PRON bracis-28409 167 2 is be AUX bracis-28409 167 3 a a DET bracis-28409 167 4 promising promising ADJ bracis-28409 167 5 result result NOUN bracis-28409 167 6 showing show VERB bracis-28409 167 7 that that SCONJ bracis-28409 167 8 the the DET bracis-28409 167 9 use use NOUN bracis-28409 167 10 of of ADP bracis-28409 167 11 a a DET bracis-28409 167 12 more more ADV bracis-28409 167 13 efficient efficient ADJ bracis-28409 167 14 selection selection NOUN bracis-28409 167 15 and and CCONJ bracis-28409 167 16 labeling labeling NOUN bracis-28409 167 17 technique technique NOUN bracis-28409 167 18 leads lead VERB bracis-28409 167 19 to to ADP bracis-28409 167 20 more more ADV bracis-28409 167 21 robust robust ADJ bracis-28409 167 22 ssl ssl ADJ bracis-28409 167 23 methods method NOUN bracis-28409 167 24 . . PUNCT bracis-28409 168 1 as as SCONJ bracis-28409 168 2 previously previously ADV bracis-28409 168 3 explained explain VERB bracis-28409 168 4 , , PUNCT bracis-28409 168 5 table table NOUN bracis-28409 168 6     SPACE bracis-28409 168 7 3 3 NUM bracis-28409 168 8 presents present VERB bracis-28409 168 9 the the DET bracis-28409 168 10 performance performance NOUN bracis-28409 168 11 of of ADP bracis-28409 168 12 the the DET bracis-28409 168 13 proposed propose VERB bracis-28409 168 14 method method NOUN bracis-28409 168 15 using use VERB bracis-28409 168 16 the the DET bracis-28409 168 17 heterogeneous heterogeneous ADJ bracis-28409 168 18 ensemble ensemble ADJ bracis-28409 168 19 flexcon flexcon NOUN bracis-28409 168 20 - - PUNCT bracis-28409 168 21 ce(het ce(het NOUN bracis-28409 168 22 ) ) PUNCT bracis-28409 168 23 , , PUNCT bracis-28409 168 24 and and CCONJ bracis-28409 168 25 the the DET bracis-28409 168 26 best good ADJ bracis-28409 168 27 results result NOUN bracis-28409 168 28 obtained obtain VERB bracis-28409 168 29 in in ADP bracis-28409 168 30 table table NOUN bracis-28409 168 31     SPACE bracis-28409 168 32 2 2 NUM bracis-28409 168 33 , , PUNCT bracis-28409 168 34 for for ADP bracis-28409 168 35 comparison comparison NOUN bracis-28409 168 36 purposes purpose NOUN bracis-28409 168 37 . . PUNCT bracis-28409 169 1 when when SCONJ bracis-28409 169 2 observing observe VERB bracis-28409 169 3 the the DET bracis-28409 169 4 aforementioned aforementioned ADJ bracis-28409 169 5 table table NOUN bracis-28409 169 6 , , PUNCT bracis-28409 169 7 it it PRON bracis-28409 169 8 is be AUX bracis-28409 169 9 possible possible ADJ bracis-28409 169 10 to to PART bracis-28409 169 11 identify identify VERB bracis-28409 169 12 that that SCONJ bracis-28409 169 13 the the DET bracis-28409 169 14 flexcon flexcon NOUN bracis-28409 169 15 - - PUNCT bracis-28409 169 16 ce(hom ce(hom ADJ bracis-28409 169 17 ) ) PUNCT bracis-28409 169 18 method method NOUN bracis-28409 169 19 obtained obtain VERB bracis-28409 169 20 the the DET bracis-28409 169 21 best good ADJ bracis-28409 169 22 performance performance NOUN bracis-28409 169 23 in in ADP bracis-28409 169 24 80 80 NUM bracis-28409 169 25 % % NOUN bracis-28409 169 26 of of ADP bracis-28409 169 27 the the DET bracis-28409 169 28 cases case NOUN bracis-28409 169 29 ( ( PUNCT bracis-28409 169 30 4 4 NUM bracis-28409 169 31 out out ADP bracis-28409 169 32 of of ADP bracis-28409 169 33 5 5 NUM bracis-28409 169 34 % % NOUN bracis-28409 169 35 of of ADP bracis-28409 169 36 initially initially ADV bracis-28409 169 37 labeled label VERB bracis-28409 169 38 instances instance NOUN bracis-28409 169 39 ) ) PUNCT bracis-28409 169 40 when when SCONJ bracis-28409 169 41 the the DET bracis-28409 169 42 number number NOUN bracis-28409 169 43 of of ADP bracis-28409 169 44 initially initially ADV bracis-28409 169 45 labeled label VERB bracis-28409 169 46 instances instance NOUN bracis-28409 169 47 was be AUX bracis-28409 169 48 higher high ADJ bracis-28409 169 49 than than ADP bracis-28409 169 50 or or CCONJ bracis-28409 169 51 equal equal ADJ bracis-28409 169 52 to to ADP bracis-28409 169 53 10 10 NUM bracis-28409 169 54 % % NOUN bracis-28409 169 55 . . PUNCT bracis-28409 170 1 on on ADP bracis-28409 170 2 the the DET bracis-28409 170 3 other other ADJ bracis-28409 170 4 hand hand NOUN bracis-28409 170 5 , , PUNCT bracis-28409 170 6 the the DET bracis-28409 170 7 other other ADJ bracis-28409 170 8 proposed propose VERB bracis-28409 170 9 method method NOUN bracis-28409 170 10 of of ADP bracis-28409 170 11 this this DET bracis-28409 170 12 paper paper NOUN bracis-28409 170 13 , , PUNCT bracis-28409 170 14 flexcon flexcon NOUN bracis-28409 170 15 - - PUNCT bracis-28409 170 16 ce(het ce(het NOUN bracis-28409 170 17 ) ) PUNCT bracis-28409 170 18 , , PUNCT bracis-28409 170 19 obtained obtain VERB bracis-28409 170 20 the the DET bracis-28409 170 21 best good ADJ bracis-28409 170 22 performance performance NOUN bracis-28409 170 23 when when SCONJ bracis-28409 170 24 the the DET bracis-28409 170 25 number number NOUN bracis-28409 170 26 of of ADP bracis-28409 170 27 initially initially ADV bracis-28409 170 28 labeled label VERB bracis-28409 170 29 instances instance NOUN bracis-28409 170 30 was be AUX bracis-28409 170 31 limited limit VERB bracis-28409 170 32 to to ADP bracis-28409 170 33 5 5 NUM bracis-28409 170 34 % % NOUN bracis-28409 170 35 . . PUNCT bracis-28409 171 1 in in ADP bracis-28409 171 2 summary summary NOUN bracis-28409 171 3 , , PUNCT bracis-28409 171 4 it it PRON bracis-28409 171 5 is be AUX bracis-28409 171 6 possible possible ADJ bracis-28409 171 7 to to PART bracis-28409 171 8 state state VERB bracis-28409 171 9 that that SCONJ bracis-28409 171 10 one one NUM bracis-28409 171 11 proposed propose VERB bracis-28409 171 12 version version NOUN bracis-28409 171 13 , , PUNCT bracis-28409 171 14 flexcon flexcon NOUN bracis-28409 171 15 - - PUNCT bracis-28409 171 16 ce(hom ce(hom ADJ bracis-28409 171 17 ) ) PUNCT bracis-28409 171 18 or or CCONJ bracis-28409 171 19 flexcon flexcon NOUN bracis-28409 171 20 - - PUNCT bracis-28409 171 21 ce(het ce(het NOUN bracis-28409 171 22 ) ) PUNCT bracis-28409 171 23 , , PUNCT bracis-28409 171 24 achieved achieve VERB bracis-28409 171 25 better well ADJ bracis-28409 171 26 accuracy accuracy NOUN bracis-28409 171 27 results result NOUN bracis-28409 171 28 than than ADP bracis-28409 171 29 all all DET bracis-28409 171 30 flexcon flexcon ADJ bracis-28409 171 31 - - PUNCT bracis-28409 171 32 c c NOUN bracis-28409 171 33 versions version NOUN bracis-28409 171 34 , , PUNCT bracis-28409 171 35 regardless regardless ADV bracis-28409 171 36 of of ADP bracis-28409 171 37 the the DET bracis-28409 171 38 type type NOUN bracis-28409 171 39 of of ADP bracis-28409 171 40 classifier classifier NOUN bracis-28409 171 41 used use VERB bracis-28409 171 42 . . PUNCT bracis-28409 172 1 the the DET bracis-28409 172 2 next next ADJ bracis-28409 172 3 subsection subsection NOUN bracis-28409 172 4 presents present VERB bracis-28409 172 5 the the DET bracis-28409 172 6 statistical statistical ADJ bracis-28409 172 7 analysis analysis NOUN bracis-28409 172 8 of of ADP bracis-28409 172 9 the the DET bracis-28409 172 10 obtained obtain VERB bracis-28409 172 11 results result NOUN bracis-28409 172 12 in in ADP bracis-28409 172 13 order order NOUN bracis-28409 172 14 to to PART bracis-28409 172 15 demonstrate demonstrate VERB bracis-28409 172 16 that that SCONJ bracis-28409 172 17 the the DET bracis-28409 172 18 proposed propose VERB bracis-28409 172 19 semi semi ADJ bracis-28409 172 20 - - ADJ bracis-28409 172 21 supervised supervised ADJ bracis-28409 172 22 machine machine NOUN bracis-28409 172 23 learning learning NOUN bracis-28409 172 24 methods method NOUN bracis-28409 172 25 are be AUX bracis-28409 172 26 promising promise VERB bracis-28409 172 27 . . PUNCT bracis-28409 173 1 6.2 6.2 NUM bracis-28409 173 2 statistical statistical ADJ bracis-28409 173 3 analysis analysis NOUN bracis-28409 173 4 the the DET bracis-28409 173 5 statistical statistical ADJ bracis-28409 173 6 analysis analysis NOUN bracis-28409 173 7 performed perform VERB bracis-28409 173 8 in in ADP bracis-28409 173 9 this this DET bracis-28409 173 10 article article NOUN bracis-28409 173 11 uses use VERB bracis-28409 173 12 critical critical ADJ bracis-28409 173 13 difference difference NOUN bracis-28409 173 14 ( ( PUNCT bracis-28409 173 15 cd cd NOUN bracis-28409 173 16 ) ) PUNCT bracis-28409 173 17 diagrams diagram NOUN bracis-28409 173 18 . . PUNCT bracis-28409 174 1 figures figure NOUN bracis-28409 174 2     SPACE bracis-28409 174 3 1 1 NUM bracis-28409 174 4 and and CCONJ bracis-28409 174 5 2 2 NUM bracis-28409 174 6 display display NOUN bracis-28409 174 7 the the DET bracis-28409 174 8 results result NOUN bracis-28409 174 9 obtained obtain VERB bracis-28409 174 10 using use VERB bracis-28409 174 11 the the DET bracis-28409 174 12 aforementioned aforementioned ADJ bracis-28409 174 13 diagrams diagram NOUN bracis-28409 174 14 . . PUNCT bracis-28409 175 1 the the DET bracis-28409 175 2 post post ADJ bracis-28409 175 3 - - ADJ bracis-28409 175 4 hoc hoc ADJ bracis-28409 175 5 friedman friedman NOUN bracis-28409 175 6 test test NOUN bracis-28409 175 7 is be AUX bracis-28409 175 8 used use VERB bracis-28409 175 9 to to PART bracis-28409 175 10 analyze analyze VERB bracis-28409 175 11 the the DET bracis-28409 175 12 results result NOUN bracis-28409 175 13 based base VERB bracis-28409 175 14 on on ADP bracis-28409 175 15 their their PRON bracis-28409 175 16 rankings ranking NOUN bracis-28409 175 17 . . PUNCT bracis-28409 176 1 additionally additionally ADV bracis-28409 176 2 , , PUNCT bracis-28409 176 3 in in ADP bracis-28409 176 4 these these DET bracis-28409 176 5 diagrams diagram NOUN bracis-28409 176 6 , , PUNCT bracis-28409 176 7 the the DET bracis-28409 176 8 methods method NOUN bracis-28409 176 9 located locate VERB bracis-28409 176 10 further far ADV bracis-28409 176 11 to to ADP bracis-28409 176 12 the the DET bracis-28409 176 13 left left NOUN bracis-28409 176 14 exhibit exhibit VERB bracis-28409 176 15 better well ADJ bracis-28409 176 16 results result NOUN bracis-28409 176 17 , , PUNCT bracis-28409 176 18 while while SCONJ bracis-28409 176 19 the the DET bracis-28409 176 20 ones one NOUN bracis-28409 176 21 on on ADP bracis-28409 176 22 the the DET bracis-28409 176 23 right right NOUN bracis-28409 176 24 presented present VERB bracis-28409 176 25 the the DET bracis-28409 176 26 poorest poor ADJ bracis-28409 176 27 performances performance NOUN bracis-28409 176 28 . . PUNCT bracis-28409 177 1 finally finally ADV bracis-28409 177 2 , , PUNCT bracis-28409 177 3 a a DET bracis-28409 177 4 method method NOUN bracis-28409 177 5 is be AUX bracis-28409 177 6 considered consider VERB bracis-28409 177 7 statistically statistically ADV bracis-28409 177 8 different different ADJ bracis-28409 177 9 from from ADP bracis-28409 177 10 another another PRON bracis-28409 177 11 when when SCONJ bracis-28409 177 12 both both PRON bracis-28409 177 13 are be AUX bracis-28409 177 14 not not PART bracis-28409 177 15 covered cover VERB bracis-28409 177 16 by by ADP bracis-28409 177 17 the the DET bracis-28409 177 18 critical critical ADJ bracis-28409 177 19 difference difference NOUN bracis-28409 177 20 line line NOUN bracis-28409 177 21 ( ( PUNCT bracis-28409 177 22 bold bold ADJ bracis-28409 177 23 horizontal horizontal ADJ bracis-28409 177 24 line line NOUN bracis-28409 177 25 in in ADP bracis-28409 177 26 the the DET bracis-28409 177 27 diagram diagram NOUN bracis-28409 177 28 ) ) PUNCT bracis-28409 177 29 . . PUNCT bracis-28409 178 1 otherwise otherwise ADV bracis-28409 178 2 , , PUNCT bracis-28409 178 3 when when SCONJ bracis-28409 178 4 this this DET bracis-28409 178 5 line line NOUN bracis-28409 178 6 covers cover VERB bracis-28409 178 7 two two NUM bracis-28409 178 8 or or CCONJ bracis-28409 178 9 more more ADJ bracis-28409 178 10 approaches approach NOUN bracis-28409 178 11 , , PUNCT bracis-28409 178 12 it it PRON bracis-28409 178 13 means mean VERB bracis-28409 178 14 that that SCONJ bracis-28409 178 15 the the DET bracis-28409 178 16 null null ADJ bracis-28409 178 17 hypothesis hypothesis NOUN bracis-28409 178 18 of of ADP bracis-28409 178 19 the the DET bracis-28409 178 20 friedman friedman PROPN bracis-28409 178 21 test test NOUN bracis-28409 178 22 can can AUX bracis-28409 178 23 not not PART bracis-28409 178 24 be be AUX bracis-28409 178 25 rejected reject VERB bracis-28409 178 26 . . PUNCT bracis-28409 179 1 figure figure NOUN bracis-28409 179 2     SPACE bracis-28409 179 3 1 1 NUM bracis-28409 179 4 presents present VERB bracis-28409 179 5 the the DET bracis-28409 179 6 critical critical ADJ bracis-28409 179 7 difference difference NOUN bracis-28409 179 8 diagram diagram NOUN bracis-28409 179 9 with with ADP bracis-28409 179 10 the the DET bracis-28409 179 11 results result NOUN bracis-28409 179 12 of of ADP bracis-28409 179 13 the the DET bracis-28409 179 14 statistical statistical ADJ bracis-28409 179 15 test test NOUN bracis-28409 179 16 of of ADP bracis-28409 179 17 the the DET bracis-28409 179 18 results result NOUN bracis-28409 179 19 obtained obtain VERB bracis-28409 179 20 in in ADP bracis-28409 179 21 table table NOUN bracis-28409 179 22     SPACE bracis-28409 179 23 2 2 NUM bracis-28409 179 24 . . PUNCT bracis-28409 180 1 the the DET bracis-28409 180 2 applied apply VERB bracis-28409 180 3 classification classification NOUN bracis-28409 180 4 algorithm algorithm NOUN bracis-28409 180 5 is be AUX bracis-28409 180 6 indicated indicate VERB bracis-28409 180 7 in in ADP bracis-28409 180 8 front front NOUN bracis-28409 180 9 of of ADP bracis-28409 180 10 the the DET bracis-28409 180 11 method method NOUN bracis-28409 180 12 name name NOUN bracis-28409 180 13 . . PUNCT bracis-28409 181 1 when when SCONJ bracis-28409 181 2 analyzing analyze VERB bracis-28409 181 3 the the DET bracis-28409 181 4 aforementioned aforementioned ADJ bracis-28409 181 5 diagram diagram NOUN bracis-28409 181 6 and and CCONJ bracis-28409 181 7 observing observe VERB bracis-28409 181 8 the the DET bracis-28409 181 9 ranking ranking NOUN bracis-28409 181 10 of of ADP bracis-28409 181 11 methods method NOUN bracis-28409 181 12 delivered deliver VERB bracis-28409 181 13 by by ADP bracis-28409 181 14 each each DET bracis-28409 181 15 classifier classifier NOUN bracis-28409 181 16 , , PUNCT bracis-28409 181 17 it it PRON bracis-28409 181 18 can can AUX bracis-28409 181 19 be be AUX bracis-28409 181 20 seen see VERB bracis-28409 181 21 that that SCONJ bracis-28409 181 22 flexcon flexcon NOUN bracis-28409 181 23 - - PUNCT bracis-28409 181 24 ce(hom)-tree ce(hom)-tree NOUN bracis-28409 181 25 was be AUX bracis-28409 181 26 the the DET bracis-28409 181 27 method method NOUN bracis-28409 181 28 that that PRON bracis-28409 181 29 reached reach VERB bracis-28409 181 30 the the DET bracis-28409 181 31 highest high ADJ bracis-28409 181 32 ranking ranking NOUN bracis-28409 181 33 , , PUNCT bracis-28409 181 34 which which PRON bracis-28409 181 35 confirms confirm VERB bracis-28409 181 36 the the DET bracis-28409 181 37 analysis analysis NOUN bracis-28409 181 38 carried carry VERB bracis-28409 181 39 out out ADP bracis-28409 181 40 in in ADP bracis-28409 181 41 the the DET bracis-28409 181 42 previous previous ADJ bracis-28409 181 43 section section NOUN bracis-28409 181 44 regarding regard VERB bracis-28409 181 45 the the DET bracis-28409 181 46 accuracy accuracy NOUN bracis-28409 181 47 of of ADP bracis-28409 181 48 these these DET bracis-28409 181 49 methods method NOUN bracis-28409 181 50 . . PUNCT bracis-28409 182 1 additionally additionally ADV bracis-28409 182 2 , , PUNCT bracis-28409 182 3 according accord VERB bracis-28409 182 4 to to ADP bracis-28409 182 5 the the DET bracis-28409 182 6 statistical statistical ADJ bracis-28409 182 7 analysis analysis NOUN bracis-28409 182 8 , , PUNCT bracis-28409 182 9 this this DET bracis-28409 182 10 method method NOUN bracis-28409 182 11 is be AUX bracis-28409 182 12 similar similar ADJ bracis-28409 182 13 to to ADP bracis-28409 182 14 the the DET bracis-28409 182 15 following follow VERB bracis-28409 182 16 four four NUM bracis-28409 182 17 methods method NOUN bracis-28409 182 18 , , PUNCT bracis-28409 182 19 in in ADP bracis-28409 182 20 which which PRON bracis-28409 182 21 two two NUM bracis-28409 182 22 of of ADP bracis-28409 182 23 them they PRON bracis-28409 182 24 applied apply VERB bracis-28409 182 25 a a DET bracis-28409 182 26 decision decision NOUN bracis-28409 182 27 tree tree NOUN bracis-28409 182 28 ( ( PUNCT bracis-28409 182 29 flexcon flexcon NOUN bracis-28409 182 30 - - PUNCT bracis-28409 182 31 c2 c2 PROPN bracis-28409 182 32 - - PUNCT bracis-28409 182 33 tree tree NOUN bracis-28409 182 34 and and CCONJ bracis-28409 182 35 flexcon flexcon NOUN bracis-28409 182 36 - - PUNCT bracis-28409 182 37 c1(v c1(v NOUN bracis-28409 182 38 ) ) PUNCT bracis-28409 182 39 ) ) PUNCT bracis-28409 182 40 and and CCONJ bracis-28409 182 41 the the DET bracis-28409 182 42 remaining remain VERB bracis-28409 182 43 two two NUM bracis-28409 182 44 methods method NOUN bracis-28409 182 45 used use VERB bracis-28409 182 46 a a DET bracis-28409 182 47 k k PROPN bracis-28409 182 48 - - PUNCT bracis-28409 182 49 nn nn ADJ bracis-28409 182 50 method method NOUN bracis-28409 182 51 ( ( PUNCT bracis-28409 182 52 flexcon flexcon NOUN bracis-28409 182 53 - - PUNCT bracis-28409 182 54 c2 c2 PROPN bracis-28409 182 55 and and CCONJ bracis-28409 182 56 flexcon flexcon NOUN bracis-28409 182 57 - - PUNCT bracis-28409 182 58 c1(v c1(v NOUN bracis-28409 182 59 ) ) PUNCT bracis-28409 182 60 ) ) PUNCT bracis-28409 182 61 . . PUNCT bracis-28409 183 1 on on ADP bracis-28409 183 2 the the DET bracis-28409 183 3 other other ADJ bracis-28409 183 4 hand hand NOUN bracis-28409 183 5 , , PUNCT bracis-28409 183 6 the the DET bracis-28409 183 7 winning win VERB bracis-28409 183 8 method method NOUN bracis-28409 183 9 is be AUX bracis-28409 183 10 statistically statistically ADV bracis-28409 183 11 superior superior ADJ bracis-28409 183 12 to to ADP bracis-28409 183 13 the the DET bracis-28409 183 14 remaining remain VERB bracis-28409 183 15 seven seven NUM bracis-28409 183 16 methods method NOUN bracis-28409 183 17 . . PUNCT bracis-28409 184 1 fig fig NOUN bracis-28409 184 2 . . PUNCT bracis-28409 185 1 1 1 X bracis-28409 185 2 . . X bracis-28409 185 3 critical critical ADJ bracis-28409 185 4 difference difference NOUN bracis-28409 185 5 diagram diagram NOUN bracis-28409 185 6 using use VERB bracis-28409 185 7 all all DET bracis-28409 185 8 methods method NOUN bracis-28409 185 9 full full ADJ bracis-28409 185 10 size size NOUN bracis-28409 185 11 image image NOUN bracis-28409 185 12 figure figure NOUN bracis-28409 185 13     SPACE bracis-28409 185 14 2 2 NUM bracis-28409 185 15 illustrates illustrate VERB bracis-28409 185 16 the the DET bracis-28409 185 17 critical critical ADJ bracis-28409 185 18 difference difference NOUN bracis-28409 185 19 diagram diagram NOUN bracis-28409 185 20 with with ADP bracis-28409 185 21 the the DET bracis-28409 185 22 results result NOUN bracis-28409 185 23 of of ADP bracis-28409 185 24 the the DET bracis-28409 185 25 statistical statistical ADJ bracis-28409 185 26 test test NOUN bracis-28409 185 27 of of ADP bracis-28409 185 28 the the DET bracis-28409 185 29 results result NOUN bracis-28409 185 30 obtained obtain VERB bracis-28409 185 31 in in ADP bracis-28409 185 32 table table NOUN bracis-28409 185 33     SPACE bracis-28409 185 34 3 3 NUM bracis-28409 185 35 . . PUNCT bracis-28409 185 36 when when SCONJ bracis-28409 185 37 analyzing analyze VERB bracis-28409 185 38 the the DET bracis-28409 185 39 rankings ranking NOUN bracis-28409 185 40 of of ADP bracis-28409 185 41 each each DET bracis-28409 185 42 method method NOUN bracis-28409 185 43 , , PUNCT bracis-28409 185 44 it it PRON bracis-28409 185 45 can can AUX bracis-28409 185 46 be be AUX bracis-28409 185 47 seen see VERB bracis-28409 185 48 that that SCONJ bracis-28409 185 49 the the DET bracis-28409 185 50 flexcon flexcon NOUN bracis-28409 185 51 - - PUNCT bracis-28409 185 52 ce(hom)-tree ce(hom)-tree NOUN bracis-28409 185 53 and and CCONJ bracis-28409 185 54 flexcon flexcon NOUN bracis-28409 185 55 - - PUNCT bracis-28409 185 56 ce(het ce(het NOUN bracis-28409 185 57 ) ) PUNCT bracis-28409 185 58 methods method NOUN bracis-28409 185 59 are be AUX bracis-28409 185 60 statistically statistically ADV bracis-28409 185 61 similar similar ADJ bracis-28409 185 62 and and CCONJ bracis-28409 185 63 both both PRON bracis-28409 185 64 achieved achieve VERB bracis-28409 185 65 the the DET bracis-28409 185 66 best good ADJ bracis-28409 185 67 results result NOUN bracis-28409 185 68 . . PUNCT bracis-28409 186 1 this this DET bracis-28409 186 2 observation observation NOUN bracis-28409 186 3 confirms confirm VERB bracis-28409 186 4 what what PRON bracis-28409 186 5 was be AUX bracis-28409 186 6 reported report VERB bracis-28409 186 7 in in ADP bracis-28409 186 8 the the DET bracis-28409 186 9 previous previous ADJ bracis-28409 186 10 section section NOUN bracis-28409 186 11 , , PUNCT bracis-28409 186 12 in in ADP bracis-28409 186 13 which which PRON bracis-28409 186 14 the the DET bracis-28409 186 15 proposed propose VERB bracis-28409 186 16 methods method NOUN bracis-28409 186 17 outperformed outperform VERB bracis-28409 186 18 the the DET bracis-28409 186 19 methods method NOUN bracis-28409 186 20 that that PRON bracis-28409 186 21 do do AUX bracis-28409 186 22 not not PART bracis-28409 186 23 use use VERB bracis-28409 186 24 classifier classifier ADJ bracis-28409 186 25 ensembles ensemble NOUN bracis-28409 186 26 . . PUNCT bracis-28409 187 1 fig fig NOUN bracis-28409 187 2 . . PUNCT bracis-28409 188 1 2 2 X bracis-28409 188 2 . . X bracis-28409 188 3 statistics statistic NOUN bracis-28409 188 4 performance performance NOUN bracis-28409 188 5 of of ADP bracis-28409 188 6 the the DET bracis-28409 188 7 best good ADJ bracis-28409 188 8 full full ADJ bracis-28409 188 9 size size NOUN bracis-28409 188 10 image image NOUN bracis-28409 188 11 in in ADP bracis-28409 188 12 summary summary NOUN bracis-28409 188 13 , , PUNCT bracis-28409 188 14 the the DET bracis-28409 188 15 results result NOUN bracis-28409 188 16 obtained obtain VERB bracis-28409 188 17 in in ADP bracis-28409 188 18 this this DET bracis-28409 188 19 paper paper NOUN bracis-28409 188 20 showed show VERB bracis-28409 188 21 significant significant ADJ bracis-28409 188 22 improvements improvement NOUN bracis-28409 188 23 of of ADP bracis-28409 188 24 the the DET bracis-28409 188 25 proposed propose VERB bracis-28409 188 26 method method NOUN bracis-28409 188 27 , , PUNCT bracis-28409 188 28 in in ADP bracis-28409 188 29 comparison comparison NOUN bracis-28409 188 30 to to ADP bracis-28409 188 31 the the DET bracis-28409 188 32 flexcon flexcon NOUN bracis-28409 188 33 - - PUNCT bracis-28409 188 34 c c NOUN bracis-28409 188 35 versions version NOUN bracis-28409 188 36 . . PUNCT bracis-28409 189 1 these these DET bracis-28409 189 2 results result NOUN bracis-28409 189 3 corroborate corroborate VERB bracis-28409 189 4 with with ADP bracis-28409 189 5 the the DET bracis-28409 189 6 results result NOUN bracis-28409 189 7 obtained obtain VERB bracis-28409 189 8 in in ADP bracis-28409 189 9 the the DET bracis-28409 189 10 previous previous ADJ bracis-28409 189 11 section section NOUN bracis-28409 189 12 , , PUNCT bracis-28409 189 13 demonstrating demonstrate VERB bracis-28409 189 14 an an DET bracis-28409 189 15 enhancement enhancement NOUN bracis-28409 189 16 in in ADP bracis-28409 189 17 the the DET bracis-28409 189 18 performance performance NOUN bracis-28409 189 19 of of ADP bracis-28409 189 20 the the DET bracis-28409 189 21 semi semi ADJ bracis-28409 189 22 - - ADJ bracis-28409 189 23 supervised supervised ADJ bracis-28409 189 24 methods method NOUN bracis-28409 189 25 of of ADP bracis-28409 189 26 the the DET bracis-28409 189 27 proposed propose VERB bracis-28409 189 28 methods method NOUN bracis-28409 189 29 . . PUNCT bracis-28409 190 1 moreover moreover ADV bracis-28409 190 2 , , PUNCT bracis-28409 190 3 the the DET bracis-28409 190 4 proposed propose VERB bracis-28409 190 5 methods method NOUN bracis-28409 190 6 are be AUX bracis-28409 190 7 validated validate VERB bracis-28409 190 8 from from ADP bracis-28409 190 9 two two NUM bracis-28409 190 10 different different ADJ bracis-28409 190 11 perspectives perspective NOUN bracis-28409 190 12 , , PUNCT bracis-28409 190 13 both both CCONJ bracis-28409 190 14 in in ADP bracis-28409 190 15 terms term NOUN bracis-28409 190 16 of of ADP bracis-28409 190 17 performance performance NOUN bracis-28409 190 18 and and CCONJ bracis-28409 190 19 statistical statistical ADJ bracis-28409 190 20 analysis analysis NOUN bracis-28409 190 21 . . PUNCT bracis-28409 191 1 7 7 NUM bracis-28409 191 2 conclusion conclusion NOUN bracis-28409 191 3 and and CCONJ bracis-28409 191 4     SPACE bracis-28409 191 5 future future NOUN bracis-28409 191 6 works work VERB bracis-28409 191 7 this this DET bracis-28409 191 8 paper paper NOUN bracis-28409 191 9 proposed propose VERB bracis-28409 191 10 a a DET bracis-28409 191 11 new new ADJ bracis-28409 191 12 ssl ssl ADJ bracis-28409 191 13 method method NOUN bracis-28409 191 14 , , PUNCT bracis-28409 191 15 called call VERB bracis-28409 191 16 flexcon flexcon NOUN bracis-28409 191 17 - - PUNCT bracis-28409 191 18 ce ce NOUN bracis-28409 191 19 , , PUNCT bracis-28409 191 20 which which PRON bracis-28409 191 21 was be AUX bracis-28409 191 22 designed design VERB bracis-28409 191 23 based base VERB bracis-28409 191 24 on on ADP bracis-28409 191 25 introducing introduce VERB bracis-28409 191 26 classifier classifier NOUN bracis-28409 191 27 ensembles ensemble NOUN bracis-28409 191 28 in in ADP bracis-28409 191 29 the the DET bracis-28409 191 30 flexcon flexcon NOUN bracis-28409 191 31 - - PUNCT bracis-28409 191 32 c c NOUN bracis-28409 191 33 functioning functioning NOUN bracis-28409 191 34 . . PUNCT bracis-28409 192 1 the the DET bracis-28409 192 2 main main ADJ bracis-28409 192 3 difference difference NOUN bracis-28409 192 4 between between ADP bracis-28409 192 5 flexcon flexcon NOUN bracis-28409 192 6 - - PUNCT bracis-28409 192 7 c c NOUN bracis-28409 192 8 and and CCONJ bracis-28409 192 9 flexcon flexcon NOUN bracis-28409 192 10 - - PUNCT bracis-28409 192 11 ce ce PROPN bracis-28409 192 12 is be AUX bracis-28409 192 13 that that SCONJ bracis-28409 192 14 the the DET bracis-28409 192 15 latter latter ADJ bracis-28409 192 16 uses use VERB bracis-28409 192 17 classifier classifier NOUN bracis-28409 192 18 ensembles ensemble NOUN bracis-28409 192 19 to to PART bracis-28409 192 20 define define VERB bracis-28409 192 21 the the DET bracis-28409 192 22 adaptive adaptive ADJ bracis-28409 192 23 confidence confidence NOUN bracis-28409 192 24 while while SCONJ bracis-28409 192 25 the the DET bracis-28409 192 26 predictions prediction NOUN bracis-28409 192 27 of of ADP bracis-28409 192 28 a a DET bracis-28409 192 29 classifier classifier NOUN bracis-28409 192 30 is be AUX bracis-28409 192 31 used use VERB bracis-28409 192 32 by by ADP bracis-28409 192 33 flexcon flexcon NOUN bracis-28409 192 34 - - PUNCT bracis-28409 192 35 c. c. NOUN bracis-28409 192 36 additionally additionally ADV bracis-28409 192 37 , , PUNCT bracis-28409 192 38 this this DET bracis-28409 192 39 work work NOUN bracis-28409 192 40 explores explore VERB bracis-28409 192 41 the the DET bracis-28409 192 42 use use NOUN bracis-28409 192 43 of of ADP bracis-28409 192 44 two two NUM bracis-28409 192 45 versions version NOUN bracis-28409 192 46 of of ADP bracis-28409 192 47 the the DET bracis-28409 192 48 proposed propose VERB bracis-28409 192 49 method method NOUN bracis-28409 192 50 : : PUNCT bracis-28409 192 51 one one NUM bracis-28409 192 52 that that PRON bracis-28409 192 53 uses use VERB bracis-28409 192 54 an an DET bracis-28409 192 55 homogeneous homogeneous ADJ bracis-28409 192 56 ensemble ensemble ADJ bracis-28409 192 57 structure structure NOUN bracis-28409 192 58 and and CCONJ bracis-28409 192 59 the the DET bracis-28409 192 60 other other ADJ bracis-28409 192 61 one one NOUN bracis-28409 192 62 used use VERB bracis-28409 192 63 an an DET bracis-28409 192 64 heterogeneous heterogeneous ADJ bracis-28409 192 65 ensemble ensemble ADJ bracis-28409 192 66 structure structure NOUN bracis-28409 192 67 . . PUNCT bracis-28409 193 1 in in ADP bracis-28409 193 2 order order NOUN bracis-28409 193 3 to to PART bracis-28409 193 4 validate validate VERB bracis-28409 193 5 the the DET bracis-28409 193 6 proposed propose VERB bracis-28409 193 7 method method NOUN bracis-28409 193 8 , , PUNCT bracis-28409 193 9 an an DET bracis-28409 193 10 empirical empirical ADJ bracis-28409 193 11 analysis analysis NOUN bracis-28409 193 12 was be AUX bracis-28409 193 13 conducted conduct VERB bracis-28409 193 14 , , PUNCT bracis-28409 193 15 in in ADP bracis-28409 193 16 which which PRON bracis-28409 193 17 a a DET bracis-28409 193 18 comparative comparative ADJ bracis-28409 193 19 analysis analysis NOUN bracis-28409 193 20 of of ADP bracis-28409 193 21 the the DET bracis-28409 193 22 performance performance NOUN bracis-28409 193 23 of of ADP bracis-28409 193 24 flexcon flexcon NOUN bracis-28409 193 25 - - PUNCT bracis-28409 193 26 ce ce NOUN bracis-28409 193 27 with with ADP bracis-28409 193 28 flexcon flexcon NOUN bracis-28409 193 29 - - PUNCT bracis-28409 193 30 c c NOUN bracis-28409 193 31 was be AUX bracis-28409 193 32 carried carry VERB bracis-28409 193 33 out out ADP bracis-28409 193 34 . . PUNCT bracis-28409 194 1 in in ADP bracis-28409 194 2 general general ADJ bracis-28409 194 3 , , PUNCT bracis-28409 194 4 the the DET bracis-28409 194 5 performance performance NOUN bracis-28409 194 6 of of ADP bracis-28409 194 7 the the DET bracis-28409 194 8 analyzed analyze VERB bracis-28409 194 9 methods method NOUN bracis-28409 194 10 was be AUX bracis-28409 194 11 assessed assess VERB bracis-28409 194 12 in in ADP bracis-28409 194 13 terms term NOUN bracis-28409 194 14 of of ADP bracis-28409 194 15 accuracy accuracy NOUN bracis-28409 194 16 and and CCONJ bracis-28409 194 17 standard standard ADJ bracis-28409 194 18 deviation deviation NOUN bracis-28409 194 19 and and CCONJ bracis-28409 194 20 they they PRON bracis-28409 194 21 were be AUX bracis-28409 194 22 validated validate VERB bracis-28409 194 23 statistically statistically ADV bracis-28409 194 24 using use VERB bracis-28409 194 25 the the DET bracis-28409 194 26 friedman friedman PROPN bracis-28409 194 27 test test NOUN bracis-28409 194 28 . . PUNCT bracis-28409 195 1 after after ADP bracis-28409 195 2 analyzing analyze VERB bracis-28409 195 3 the the DET bracis-28409 195 4 results result NOUN bracis-28409 195 5 , , PUNCT bracis-28409 195 6 it it PRON bracis-28409 195 7 was be AUX bracis-28409 195 8 concluded conclude VERB bracis-28409 195 9 that that SCONJ bracis-28409 195 10 the the DET bracis-28409 195 11 flexcon flexcon NOUN bracis-28409 195 12 - - PUNCT bracis-28409 195 13 ce ce PROPN bracis-28409 195 14 method method NOUN bracis-28409 195 15 obtained obtain VERB bracis-28409 195 16 better well ADJ bracis-28409 195 17 results result NOUN bracis-28409 195 18 in in ADP bracis-28409 195 19 most most ADJ bracis-28409 195 20 cases case NOUN bracis-28409 195 21 , , PUNCT bracis-28409 195 22 with with ADP bracis-28409 195 23 emphasis emphasis NOUN bracis-28409 195 24 on on ADP bracis-28409 195 25 the the DET bracis-28409 195 26 version version NOUN bracis-28409 195 27 that that PRON bracis-28409 195 28 used use VERB bracis-28409 195 29 the the DET bracis-28409 195 30 homogeneous homogeneous ADJ bracis-28409 195 31 ensemble ensemble NOUN bracis-28409 195 32 with with ADP bracis-28409 195 33 the the DET bracis-28409 195 34 tree tree NOUN bracis-28409 195 35 classifier classifier NOUN bracis-28409 195 36 , , PUNCT bracis-28409 195 37 followed follow VERB bracis-28409 195 38 by by ADP bracis-28409 195 39 the the DET bracis-28409 195 40 heterogeneous heterogeneous ADJ bracis-28409 195 41 flexcon flexcon NOUN bracis-28409 195 42 - - PUNCT bracis-28409 195 43 ce ce NOUN bracis-28409 195 44 version version NOUN bracis-28409 195 45 . . PUNCT bracis-28409 196 1 for for ADP bracis-28409 196 2 future future ADJ bracis-28409 196 3 work work NOUN bracis-28409 196 4 , , PUNCT bracis-28409 196 5 we we PRON bracis-28409 196 6 suggest suggest VERB bracis-28409 196 7 to to PART bracis-28409 196 8 explore explore VERB bracis-28409 196 9 other other ADJ bracis-28409 196 10 classification classification NOUN bracis-28409 196 11 algorithms algorithm NOUN bracis-28409 196 12 and and CCONJ bracis-28409 196 13 use use VERB bracis-28409 196 14 new new ADJ bracis-28409 196 15 databases database NOUN bracis-28409 196 16 to to PART bracis-28409 196 17 address address VERB bracis-28409 196 18 experimental experimental ADJ bracis-28409 196 19 scenarios scenario NOUN bracis-28409 196 20 . . PUNCT bracis-28409 197 1 furthermore furthermore ADV bracis-28409 197 2 , , PUNCT bracis-28409 197 3 it it PRON bracis-28409 197 4 is be AUX bracis-28409 197 5 recommended recommend VERB bracis-28409 197 6 to to PART bracis-28409 197 7 make make VERB bracis-28409 197 8 the the DET bracis-28409 197 9 size size NOUN bracis-28409 197 10 of of ADP bracis-28409 197 11 the the DET bracis-28409 197 12 classifier classifier NOUN bracis-28409 197 13 ensembles ensemble VERB bracis-28409 197 14 more more ADV bracis-28409 197 15 flexible flexible ADJ bracis-28409 197 16 in in ADP bracis-28409 197 17 relation relation NOUN bracis-28409 197 18 to to ADP bracis-28409 197 19 the the DET bracis-28409 197 20 number number NOUN bracis-28409 197 21 of of ADP bracis-28409 197 22 base base NOUN bracis-28409 197 23 classifiers classifier NOUN bracis-28409 197 24 . . PUNCT bracis-28409 198 1 references reference NOUN bracis-28409 198 2 faceli faceli ADV bracis-28409 198 3 , , PUNCT bracis-28409 198 4 k. k. PROPN bracis-28409 198 5 , , PUNCT bracis-28409 198 6 lorena lorena PROPN bracis-28409 198 7 , , PUNCT bracis-28409 198 8 a.c a.c PROPN bracis-28409 198 9 . . PROPN bracis-28409 198 10 , , PUNCT bracis-28409 198 11 gama gama PROPN bracis-28409 198 12 , , PUNCT bracis-28409 198 13 j. j. PROPN bracis-28409 198 14 , , PUNCT bracis-28409 198 15 de de PROPN bracis-28409 198 16 leon leon PROPN bracis-28409 198 17 ferreira ferreira PROPN bracis-28409 198 18 de de PROPN bracis-28409 198 19 carvalho carvalho PROPN bracis-28409 198 20 , , PUNCT bracis-28409 198 21 a.c.p a.c.p PROPN bracis-28409 198 22 . . PUNCT bracis-28409 198 23 : : PUNCT bracis-28409 198 24 an an DET bracis-28409 198 25 approach approach NOUN bracis-28409 198 26 of of ADP bracis-28409 198 27 machine machine NOUN bracis-28409 198 28 learning learning NOUN bracis-28409 198 29 . . PUNCT bracis-28409 199 1 artif artif INTJ bracis-28409 199 2 . . PUNCT bracis-28409 200 1 intell intell PROPN bracis-28409 200 2 . . PUNCT bracis-28409 201 1 ( ( PUNCT bracis-28409 201 2 2011 2011 NUM bracis-28409 201 3 ) ) PUNCT bracis-28409 201 4 google google PROPN bracis-28409 201 5 scholar scholar NOUN bracis-28409 201 6     SPACE bracis-28409 201 7 chapelle chapelle NOUN bracis-28409 201 8 , , PUNCT bracis-28409 201 9 o. o. PROPN bracis-28409 201 10 , , PUNCT bracis-28409 201 11 scholkopf scholkopf PROPN bracis-28409 201 12 , , PUNCT bracis-28409 201 13 b. b. PROPN bracis-28409 201 14 , , PUNCT bracis-28409 201 15 zien zien PROPN bracis-28409 201 16 , , PUNCT bracis-28409 201 17 a. a. NOUN bracis-28409 201 18 : : PUNCT bracis-28409 201 19 semi semi ADJ bracis-28409 201 20 - - ADJ bracis-28409 201 21 supervised supervised ADJ bracis-28409 201 22 learning learning NOUN bracis-28409 201 23 , , PUNCT bracis-28409 201 24 vol vol NOUN bracis-28409 201 25 . . PROPN bracis-28409 202 1 2 2 NUM bracis-28409 202 2 . . X bracis-28409 202 3 the the DET bracis-28409 202 4 mit mit PROPN bracis-28409 202 5 press press PROPN bracis-28409 202 6 , , PUNCT bracis-28409 202 7 cambridge cambridge PROPN bracis-28409 202 8 , , PUNCT bracis-28409 202 9 ma ma PROPN bracis-28409 202 10 ( ( PUNCT bracis-28409 202 11 2006 2006 NUM bracis-28409 202 12 ) ) PUNCT bracis-28409 202 13 google google PROPN bracis-28409 202 14 scholar scholar NOUN bracis-28409 202 15     SPACE bracis-28409 202 16 ovidio ovidio ADJ bracis-28409 202 17 vale vale NOUN bracis-28409 202 18 , , PUNCT bracis-28409 202 19 k.m k.m PROPN bracis-28409 202 20 . . PROPN bracis-28409 202 21 , , PUNCT bracis-28409 202 22 et et PROPN bracis-28409 202 23 al al PROPN bracis-28409 202 24 . . PUNCT bracis-28409 202 25 : : PUNCT bracis-28409 202 26 automatic automatic ADJ bracis-28409 202 27 adjustment adjustment NOUN bracis-28409 202 28 of of ADP bracis-28409 202 29 confidence confidence NOUN bracis-28409 202 30 values value NOUN bracis-28409 202 31 in in ADP bracis-28409 202 32 self self NOUN bracis-28409 202 33 - - PUNCT bracis-28409 202 34 training training NOUN bracis-28409 202 35 semi semi ADJ bracis-28409 202 36 - - ADJ bracis-28409 202 37 supervised supervised ADJ bracis-28409 202 38 method method NOUN bracis-28409 202 39 . . PUNCT bracis-28409 203 1 in in ADP bracis-28409 203 2 : : PUNCT bracis-28409 203 3 2018 2018 NUM bracis-28409 203 4 international international ADJ bracis-28409 203 5 joint joint ADJ bracis-28409 203 6 conference conference NOUN bracis-28409 203 7 on on ADP bracis-28409 203 8 neural neural ADJ bracis-28409 203 9 networks network NOUN bracis-28409 203 10 ( ( PUNCT bracis-28409 203 11 ijcnn ijcnn PROPN bracis-28409 203 12 ) ) PUNCT bracis-28409 203 13 , , PUNCT bracis-28409 203 14 pp pp ADP bracis-28409 203 15 . . PUNCT bracis-28409 204 1 1–8 1–8 X bracis-28409 204 2 . . PUNCT bracis-28409 204 3 ieee ieee PROPN bracis-28409 204 4 ( ( PUNCT bracis-28409 204 5 2018 2018 NUM bracis-28409 204 6 ) ) PUNCT bracis-28409 204 7 google google PROPN bracis-28409 204 8 scholar scholar NOUN bracis-28409 204 9     SPACE bracis-28409 204 10 wei wei PROPN bracis-28409 204 11 , , PUNCT bracis-28409 204 12 w. w. PROPN bracis-28409 204 13 , , PUNCT bracis-28409 204 14 jiang jiang PROPN bracis-28409 204 15 , , PUNCT bracis-28409 204 16 f. f. PROPN bracis-28409 204 17 , , PUNCT bracis-28409 204 18 yu yu PROPN bracis-28409 204 19 , , PUNCT bracis-28409 204 20 x. x. PROPN bracis-28409 204 21 , , PUNCT bracis-28409 204 22 du du PROPN bracis-28409 204 23 , , PUNCT bracis-28409 204 24 j. j. PROPN bracis-28409 204 25 : : PUNCT bracis-28409 204 26 an an DET bracis-28409 204 27 ensemble ensemble ADJ bracis-28409 204 28 learning learning NOUN bracis-28409 204 29 algorithm algorithm NOUN bracis-28409 204 30 based base VERB bracis-28409 204 31 on on ADP bracis-28409 204 32 resampling resample VERB bracis-28409 204 33 and and CCONJ bracis-28409 204 34 hybrid hybrid ADJ bracis-28409 204 35 feature feature NOUN bracis-28409 204 36 selection selection NOUN bracis-28409 204 37 , , PUNCT bracis-28409 204 38 with with ADP bracis-28409 204 39 an an DET bracis-28409 204 40 application application NOUN bracis-28409 204 41 to to PART bracis-28409 204 42 software software NOUN bracis-28409 204 43 defect defect NOUN bracis-28409 204 44 prediction prediction NOUN bracis-28409 204 45 . . PUNCT bracis-28409 205 1 in in ADP bracis-28409 205 2 : : PUNCT bracis-28409 205 3 2022 2022 NUM bracis-28409 205 4 7th 7th ADJ bracis-28409 205 5 international international ADJ bracis-28409 205 6 conference conference NOUN bracis-28409 205 7 on on ADP bracis-28409 205 8 information information NOUN bracis-28409 205 9 and and CCONJ bracis-28409 205 10 network network NOUN bracis-28409 205 11 technologies technology NOUN bracis-28409 205 12 ( ( PUNCT bracis-28409 205 13 icint icint PROPN bracis-28409 205 14 ) ) PUNCT bracis-28409 205 15 , , PUNCT bracis-28409 205 16 pp pp ADP bracis-28409 205 17 . . PUNCT bracis-28409 206 1 52–56 52–56 NUM bracis-28409 206 2 . . PUNCT bracis-28409 206 3 ieee ieee PROPN bracis-28409 206 4 ( ( PUNCT bracis-28409 206 5 2022 2022 NUM bracis-28409 206 6 ) ) PUNCT bracis-28409 206 7 google google PROPN bracis-28409 206 8 scholar scholar NOUN bracis-28409 206 9     SPACE bracis-28409 206 10 safiya safiya NOUN bracis-28409 206 11 parvin parvin PROPN bracis-28409 206 12 , , PUNCT bracis-28409 206 13 a. a. PROPN bracis-28409 206 14 , , PUNCT bracis-28409 206 15 saleena saleena PROPN bracis-28409 206 16 , , PUNCT bracis-28409 206 17 b. b. PROPN bracis-28409 206 18 : : PUNCT bracis-28409 206 19 an an DET bracis-28409 206 20 ensemble ensemble ADJ bracis-28409 206 21 classifier classifier NOUN bracis-28409 206 22 model model NOUN bracis-28409 206 23 to to PART bracis-28409 206 24 predict predict VERB bracis-28409 206 25 credit credit NOUN bracis-28409 206 26 scoring scoring NOUN bracis-28409 206 27 - - PUNCT bracis-28409 206 28 comparative comparative ADJ bracis-28409 206 29 analysis analysis NOUN bracis-28409 206 30 . . PUNCT bracis-28409 207 1 in in ADP bracis-28409 207 2 : : PUNCT bracis-28409 207 3 2020 2020 NUM bracis-28409 207 4 ieee ieee PROPN bracis-28409 207 5 international international ADJ bracis-28409 207 6 symposium symposium NOUN bracis-28409 207 7 on on ADP bracis-28409 207 8 smart smart ADJ bracis-28409 207 9 electronic electronic ADJ bracis-28409 207 10 systems system NOUN bracis-28409 207 11 ( ( PUNCT bracis-28409 207 12 ises ise NOUN bracis-28409 207 13 ) ) PUNCT bracis-28409 207 14 ( ( PUNCT bracis-28409 207 15 formerly formerly ADV bracis-28409 207 16 inis inis PROPN bracis-28409 207 17 ) ) PUNCT bracis-28409 207 18 , , PUNCT bracis-28409 207 19 pp pp ADJ bracis-28409 207 20 . . PUNCT bracis-28409 208 1 27–30 27–30 NUM bracis-28409 208 2 . . PUNCT bracis-28409 209 1 ieee ieee NOUN bracis-28409 209 2 ( ( PUNCT bracis-28409 209 3 2020 2020 NUM bracis-28409 209 4 ) ) PUNCT bracis-28409 209 5 google google PROPN bracis-28409 209 6 scholar scholar NOUN bracis-28409 209 7     SPACE bracis-28409 209 8 zohaib zohaib PROPN bracis-28409 209 9 jan jan PROPN bracis-28409 209 10 , , PUNCT bracis-28409 209 11 m. m. NOUN bracis-28409 209 12 , , PUNCT bracis-28409 209 13 verma verma PROPN bracis-28409 209 14 , , PUNCT bracis-28409 209 15 b. b. PROPN bracis-28409 209 16 : : PUNCT bracis-28409 209 17 a a DET bracis-28409 209 18 novel novel ADJ bracis-28409 209 19 diversity diversity NOUN bracis-28409 209 20 measure measure NOUN bracis-28409 209 21 and and CCONJ bracis-28409 209 22 classifier classifier NOUN bracis-28409 209 23 selection selection NOUN bracis-28409 209 24 approach approach NOUN bracis-28409 209 25 for for ADP bracis-28409 209 26 generating generate VERB bracis-28409 209 27 ensemble ensemble ADJ bracis-28409 209 28 classifiers classifier NOUN bracis-28409 209 29 . . PUNCT bracis-28409 210 1 ieee ieee NOUN bracis-28409 210 2 access access NOUN bracis-28409 210 3 7 7 NUM bracis-28409 210 4 , , PUNCT bracis-28409 210 5 156360–156373 156360–156373 NUM bracis-28409 210 6 ( ( PUNCT bracis-28409 210 7 2019 2019 NUM bracis-28409 210 8 ) ) PUNCT bracis-28409 210 9 google google NOUN bracis-28409 210 10 scholar scholar NOUN bracis-28409 210 11     SPACE bracis-28409 210 12 lochter lochter NOUN bracis-28409 210 13 , , PUNCT bracis-28409 210 14 j.v j.v PROPN bracis-28409 210 15 . . PROPN bracis-28409 210 16 , , PUNCT bracis-28409 210 17 zanetti zanetti PROPN bracis-28409 210 18 , , PUNCT bracis-28409 210 19 r.f r.f PROPN bracis-28409 210 20 . . PROPN bracis-28409 210 21 , , PUNCT bracis-28409 210 22 reller reller NOUN bracis-28409 210 23 , , PUNCT bracis-28409 210 24 d. d. PROPN bracis-28409 210 25 , , PUNCT bracis-28409 210 26 almeida almeida PROPN bracis-28409 210 27 , , PUNCT bracis-28409 210 28 t.a t.a PROPN bracis-28409 210 29 . . PROPN bracis-28409 210 30 : : PUNCT bracis-28409 211 1 short short ADJ bracis-28409 211 2 text text NOUN bracis-28409 211 3 opinion opinion NOUN bracis-28409 211 4 detection detection NOUN bracis-28409 211 5 using use VERB bracis-28409 211 6 ensemble ensemble NOUN bracis-28409 211 7 of of ADP bracis-28409 211 8 classifiers classifier NOUN bracis-28409 211 9 and and CCONJ bracis-28409 211 10 semantic semantic ADJ bracis-28409 211 11 indexing indexing NOUN bracis-28409 211 12 . . PUNCT bracis-28409 212 1 exp exp PROPN bracis-28409 212 2 . . PUNCT bracis-28409 212 3 syst syst PROPN bracis-28409 212 4 . . PUNCT bracis-28409 213 1 appl appl PROPN bracis-28409 213 2 . . PUNCT bracis-28409 214 1 62:243–249 62:243–249 NUM bracis-28409 214 2 ( ( PUNCT bracis-28409 214 3 2016 2016 NUM bracis-28409 214 4 ) ) PUNCT bracis-28409 214 5 google google PROPN bracis-28409 214 6 scholar scholar NOUN bracis-28409 214 7     SPACE bracis-28409 214 8 li li PROPN bracis-28409 214 9 , , PUNCT bracis-28409 214 10 x. x. PROPN bracis-28409 214 11 , , PUNCT bracis-28409 214 12 shi shi PROPN bracis-28409 214 13 , , PUNCT bracis-28409 214 14 t. t. PROPN bracis-28409 214 15 , , PUNCT bracis-28409 214 16 li li PROPN bracis-28409 214 17 , , PUNCT bracis-28409 214 18 p. p. PROPN bracis-28409 214 19 , , PUNCT bracis-28409 214 20 zhou zhou PROPN bracis-28409 214 21 , , PUNCT bracis-28409 214 22 w. w. PROPN bracis-28409 214 23 : : PUNCT bracis-28409 214 24 application application NOUN bracis-28409 214 25 of of ADP bracis-28409 214 26 bagging bag VERB bracis-28409 214 27 ensemble ensemble ADJ bracis-28409 214 28 classifier classifier NOUN bracis-28409 214 29 based base VERB bracis-28409 214 30 on on ADP bracis-28409 214 31 genetic genetic ADJ bracis-28409 214 32 algorithm algorithm NOUN bracis-28409 214 33 in in ADP bracis-28409 214 34 the the DET bracis-28409 214 35 text text NOUN bracis-28409 214 36 classification classification NOUN bracis-28409 214 37 of of ADP bracis-28409 214 38 railway railway NOUN bracis-28409 214 39 fault fault NOUN bracis-28409 214 40 hazards hazard NOUN bracis-28409 214 41 . . PUNCT bracis-28409 215 1 in in ADP bracis-28409 215 2 : : PUNCT bracis-28409 215 3 2019 2019 NUM bracis-28409 215 4 2nd 2nd ADJ bracis-28409 215 5 international international ADJ bracis-28409 215 6 conference conference NOUN bracis-28409 215 7 on on ADP bracis-28409 215 8 artificial artificial ADJ bracis-28409 215 9 intelligence intelligence NOUN bracis-28409 215 10 and and CCONJ bracis-28409 215 11 big big ADJ bracis-28409 215 12 data datum NOUN bracis-28409 215 13 ( ( PUNCT bracis-28409 215 14 icaibd icaibd NOUN bracis-28409 215 15 ) ) PUNCT bracis-28409 215 16 , , PUNCT bracis-28409 215 17 pp pp ADJ bracis-28409 215 18 . . PUNCT bracis-28409 216 1 286–290 286–290 NUM bracis-28409 216 2 . . PUNCT bracis-28409 216 3 ieee ieee NOUN bracis-28409 216 4 ( ( PUNCT bracis-28409 216 5 2019 2019 NUM bracis-28409 216 6 ) ) PUNCT bracis-28409 216 7 google google NOUN bracis-28409 216 8 scholar scholar NOUN bracis-28409 216 9     SPACE bracis-28409 216 10 cichosz cichosz NOUN bracis-28409 216 11 , , PUNCT bracis-28409 216 12 p. p. NOUN bracis-28409 216 13 : : PUNCT bracis-28409 216 14 data datum NOUN bracis-28409 216 15 mining mining NOUN bracis-28409 216 16 algorithms algorithm NOUN bracis-28409 216 17 : : PUNCT bracis-28409 216 18 explained explain VERB bracis-28409 216 19 using use VERB bracis-28409 216 20 r. r. PROPN bracis-28409 216 21 john john PROPN bracis-28409 216 22 wiley wiley PROPN bracis-28409 216 23 & & CCONJ bracis-28409 216 24 sons sons PROPN bracis-28409 216 25 ( ( PUNCT bracis-28409 216 26 2014 2014 NUM bracis-28409 216 27 ) ) PUNCT bracis-28409 216 28 google google PROPN bracis-28409 216 29 scholar scholar NOUN bracis-28409 216 30     SPACE bracis-28409 216 31 albalate albalate NOUN bracis-28409 216 32 , , PUNCT bracis-28409 216 33 a. a. NOUN bracis-28409 216 34 , , PUNCT bracis-28409 216 35 minker minker PROPN bracis-28409 216 36 , , PUNCT bracis-28409 216 37 w. w. NOUN bracis-28409 216 38 : : PUNCT bracis-28409 216 39 semi semi ADJ bracis-28409 216 40 - - ADJ bracis-28409 216 41 supervised supervised ADJ bracis-28409 216 42 and and CCONJ bracis-28409 216 43 unsupervised unsupervised ADJ bracis-28409 216 44 machine machine NOUN bracis-28409 216 45 learning learning NOUN bracis-28409 216 46 : : PUNCT bracis-28409 216 47 novel novel ADJ bracis-28409 216 48 strategies strategy NOUN bracis-28409 216 49 . . PUNCT bracis-28409 217 1 john john PROPN bracis-28409 217 2 wiley wiley PROPN bracis-28409 217 3 & & CCONJ bracis-28409 217 4 sons sons PROPN bracis-28409 217 5 ( ( PUNCT bracis-28409 217 6 2013 2013 NUM bracis-28409 217 7 ) ) PUNCT bracis-28409 217 8 google google PROPN bracis-28409 217 9 scholar scholar NOUN bracis-28409 217 10     SPACE bracis-28409 217 11 zhu zhu PROPN bracis-28409 217 12 , , PUNCT bracis-28409 217 13 x. x. PROPN bracis-28409 217 14 , , PUNCT bracis-28409 217 15 goldberg goldberg PROPN bracis-28409 217 16 , , PUNCT bracis-28409 217 17 a.b a.b PROPN bracis-28409 217 18 . . PROPN bracis-28409 217 19 : : PUNCT bracis-28409 217 20 introduction introduction NOUN bracis-28409 217 21 to to ADP bracis-28409 217 22 semi semi ADJ bracis-28409 217 23 - - ADJ bracis-28409 217 24 supervised supervised ADJ bracis-28409 217 25 learning learning NOUN bracis-28409 217 26 . . PUNCT bracis-28409 218 1 synth synth ADJ bracis-28409 218 2 . . PUNCT bracis-28409 219 1 lect lect PROPN bracis-28409 219 2 . . PUNCT bracis-28409 220 1 artif artif INTJ bracis-28409 220 2 . . PUNCT bracis-28409 221 1 intell intell PROPN bracis-28409 221 2 . . PUNCT bracis-28409 222 1 mach mach PROPN bracis-28409 222 2 . . PUNCT bracis-28409 223 1 learn learn VERB bracis-28409 223 2 . . PUNCT bracis-28409 224 1 3(1 3(1 NUM bracis-28409 224 2 ) ) PUNCT bracis-28409 224 3 , , PUNCT bracis-28409 224 4 1–130 1–130 NUM bracis-28409 224 5 ( ( PUNCT bracis-28409 224 6 2009 2009 NUM bracis-28409 224 7 ) ) PUNCT bracis-28409 224 8 math math NOUN bracis-28409 224 9     SPACE bracis-28409 224 10 google google PROPN bracis-28409 224 11 scholar scholar NOUN bracis-28409 224 12     SPACE bracis-28409 224 13 rodrigues rodrigue NOUN bracis-28409 224 14 , , PUNCT bracis-28409 224 15 f.m f.m PROPN bracis-28409 224 16 . . PROPN bracis-28409 224 17 , , PUNCT bracis-28409 224 18 de de PROPN bracis-28409 224 19 m. m. NOUN bracis-28409 224 20 santos santos PROPN bracis-28409 224 21 , , PUNCT bracis-28409 224 22 a. a. NOUN bracis-28409 224 23 , , PUNCT bracis-28409 224 24 canuto canuto NOUN bracis-28409 224 25 , , PUNCT bracis-28409 224 26 a.m.p a.m.p PROPN bracis-28409 224 27 . . PUNCT bracis-28409 224 28 : : PUNCT bracis-28409 224 29 using use VERB bracis-28409 224 30 confidence confidence NOUN bracis-28409 224 31 values value NOUN bracis-28409 224 32 in in ADP bracis-28409 224 33 multi multi ADJ bracis-28409 224 34 - - ADJ bracis-28409 224 35 label label ADJ bracis-28409 224 36 classification classification NOUN bracis-28409 224 37 problems problem NOUN bracis-28409 224 38 with with ADP bracis-28409 224 39 semi semi ADJ bracis-28409 224 40 - - ADJ bracis-28409 224 41 supervised supervised ADJ bracis-28409 224 42 learning learning NOUN bracis-28409 224 43 . . PUNCT bracis-28409 225 1 in in ADP bracis-28409 225 2 : : PUNCT bracis-28409 225 3 the the DET bracis-28409 225 4 2013 2013 NUM bracis-28409 225 5 international international ADJ bracis-28409 225 6 joint joint ADJ bracis-28409 225 7 conference conference NOUN bracis-28409 225 8 on on ADP bracis-28409 225 9 neural neural ADJ bracis-28409 225 10 networks network NOUN bracis-28409 225 11 ( ( PUNCT bracis-28409 225 12 ijcnn ijcnn PROPN bracis-28409 225 13 ) ) PUNCT bracis-28409 225 14 , , PUNCT bracis-28409 225 15 pp pp ADP bracis-28409 225 16 . . PUNCT bracis-28409 226 1 1–8 1–8 X bracis-28409 226 2 . . PUNCT bracis-28409 226 3 ieee ieee PROPN bracis-28409 226 4 ( ( PUNCT bracis-28409 226 5 2013 2013 NUM bracis-28409 226 6 ) ) PUNCT bracis-28409 226 7 google google PROPN bracis-28409 226 8 scholar scholar NOUN bracis-28409 226 9     SPACE bracis-28409 226 10 nascimento nascimento PROPN bracis-28409 226 11 , , PUNCT bracis-28409 226 12 d.s.c d.s.c PROPN bracis-28409 226 13 . . PROPN bracis-28409 226 14 , , PUNCT bracis-28409 226 15 coelho coelho PROPN bracis-28409 226 16 , , PUNCT bracis-28409 226 17 a.l.v a.l.v PROPN bracis-28409 226 18 . . PROPN bracis-28409 226 19 , , PUNCT bracis-28409 226 20 canuto canuto PROPN bracis-28409 226 21 , , PUNCT bracis-28409 226 22 a.m.p a.m.p PROPN bracis-28409 226 23 . . PUNCT bracis-28409 226 24 : : PUNCT bracis-28409 226 25 integrating integrate VERB bracis-28409 226 26 complementary complementary ADJ bracis-28409 226 27 techniques technique NOUN bracis-28409 226 28 for for ADP bracis-28409 226 29 promoting promote VERB bracis-28409 226 30 diversity diversity NOUN bracis-28409 226 31 in in ADP bracis-28409 226 32 classifier classifier ADJ bracis-28409 226 33 ensembles ensemble NOUN bracis-28409 226 34 : : PUNCT bracis-28409 226 35 a a DET bracis-28409 226 36 systematic systematic ADJ bracis-28409 226 37 study study NOUN bracis-28409 226 38 . . PUNCT bracis-28409 227 1 neurocomputing neurocompute VERB bracis-28409 227 2 138 138 NUM bracis-28409 227 3 , , PUNCT bracis-28409 227 4 347–357 347–357 NUM bracis-28409 227 5 ( ( PUNCT bracis-28409 227 6 2014 2014 NUM bracis-28409 227 7 ) ) PUNCT bracis-28409 227 8 article article NOUN bracis-28409 227 9     SPACE bracis-28409 227 10 google google PROPN bracis-28409 227 11 scholar scholar NOUN bracis-28409 227 12     SPACE bracis-28409 227 13 gharroudi gharroudi NOUN bracis-28409 227 14 , , PUNCT bracis-28409 227 15 o. o. NOUN bracis-28409 227 16 : : PUNCT bracis-28409 227 17 ensemble ensemble ADJ bracis-28409 227 18 multi multi ADJ bracis-28409 227 19 - - ADJ bracis-28409 227 20 label label ADJ bracis-28409 227 21 learning learning NOUN bracis-28409 227 22 in in ADP bracis-28409 227 23 supervised supervised ADJ bracis-28409 227 24 and and CCONJ bracis-28409 227 25 semi semi ADJ bracis-28409 227 26 - - ADJ bracis-28409 227 27 supervised supervised ADJ bracis-28409 227 28 settings setting NOUN bracis-28409 227 29 . . PUNCT bracis-28409 228 1 ph.d ph.d PROPN bracis-28409 228 2 . . PUNCT bracis-28409 229 1 thesis thesis PROPN bracis-28409 229 2 , , PUNCT bracis-28409 229 3 université université PROPN bracis-28409 229 4 de de PROPN bracis-28409 229 5 lyon lyon PROPN bracis-28409 229 6 ( ( PUNCT bracis-28409 229 7 2017 2017 NUM bracis-28409 229 8 ) ) PUNCT bracis-28409 229 9 google google PROPN bracis-28409 229 10 scholar scholar NOUN bracis-28409 229 11     SPACE bracis-28409 229 12 rodrigues rodrigue NOUN bracis-28409 229 13 , , PUNCT bracis-28409 229 14 f.m f.m PROPN bracis-28409 229 15 . . PROPN bracis-28409 229 16 , , PUNCT bracis-28409 229 17 câmara câmara PROPN bracis-28409 229 18 , , PUNCT bracis-28409 229 19 c.j c.j PROPN bracis-28409 229 20 . . PROPN bracis-28409 229 21 , , PUNCT bracis-28409 229 22 canuto canuto PROPN bracis-28409 229 23 , , PUNCT bracis-28409 229 24 a.m.p a.m.p PROPN bracis-28409 229 25 . . PROPN bracis-28409 229 26 , , PUNCT bracis-28409 229 27 santos santos PROPN bracis-28409 229 28 , , PUNCT bracis-28409 229 29 a.m. a.m. ADV bracis-28409 229 30 : : PUNCT bracis-28409 229 31 confidence confidence NOUN bracis-28409 229 32 factor factor NOUN bracis-28409 229 33 and and CCONJ bracis-28409 229 34 feature feature NOUN bracis-28409 229 35 selection selection NOUN bracis-28409 229 36 for for ADP bracis-28409 229 37 semi semi ADJ bracis-28409 229 38 - - ADJ bracis-28409 229 39 supervised supervised ADJ bracis-28409 229 40 multi multi ADJ bracis-28409 229 41 - - ADJ bracis-28409 229 42 label label ADJ bracis-28409 229 43 classification classification NOUN bracis-28409 229 44 methods method NOUN bracis-28409 229 45 . . PUNCT bracis-28409 230 1 in in ADP bracis-28409 230 2 : : PUNCT bracis-28409 230 3 2014 2014 NUM bracis-28409 230 4 international international ADJ bracis-28409 230 5 joint joint ADJ bracis-28409 230 6 conference conference NOUN bracis-28409 230 7 on on ADP bracis-28409 230 8 neural neural ADJ bracis-28409 230 9 networks network NOUN bracis-28409 230 10 ( ( PUNCT bracis-28409 230 11 ijcnn ijcnn PROPN bracis-28409 230 12 ) ) PUNCT bracis-28409 230 13 , , PUNCT bracis-28409 230 14 pp pp ADP bracis-28409 230 15 . . PUNCT bracis-28409 231 1 864–871 864–871 NUM bracis-28409 231 2 . . PUNCT bracis-28409 232 1 ieee ieee NOUN bracis-28409 232 2 ( ( PUNCT bracis-28409 232 3 2014 2014 NUM bracis-28409 232 4 ) ) PUNCT bracis-28409 232 5 google google PROPN bracis-28409 232 6 scholar scholar NOUN bracis-28409 232 7     SPACE bracis-28409 232 8 gorgônio gorgônio NOUN bracis-28409 232 9 , , PUNCT bracis-28409 232 10 a.c a.c PROPN bracis-28409 232 11 . . PROPN bracis-28409 232 12 , , PUNCT bracis-28409 232 13 et et PROPN bracis-28409 232 14 al al PROPN bracis-28409 232 15 . . PUNCT bracis-28409 232 16 : : PUNCT bracis-28409 232 17 análise análise PROPN bracis-28409 232 18 da da PROPN bracis-28409 232 19 variação variação PROPN bracis-28409 232 20 do do AUX bracis-28409 232 21 limiar limiar VERB bracis-28409 232 22 para para PROPN bracis-28409 232 23 o o PROPN bracis-28409 232 24 algoritmo algoritmo X bracis-28409 232 25 de de X bracis-28409 232 26 aprendizado aprendizado PROPN bracis-28409 232 27 semissupervisionado semissupervisionado PROPN bracis-28409 232 28 flexcon flexcon PROPN bracis-28409 232 29 - - PUNCT bracis-28409 232 30 c c NOUN bracis-28409 232 31 / / SYM bracis-28409 232 32 threshold threshold NOUN bracis-28409 232 33 variation variation NOUN bracis-28409 232 34 analysis analysis NOUN bracis-28409 232 35 for for ADP bracis-28409 232 36 flexcon flexcon NOUN bracis-28409 232 37 - - PUNCT bracis-28409 232 38 c c NOUN bracis-28409 232 39 semisupervised semisupervise VERB bracis-28409 232 40 learning learning NOUN bracis-28409 232 41 algorithm algorithm NOUN bracis-28409 232 42 . . PUNCT bracis-28409 233 1 brazil brazil PROPN bracis-28409 233 2 . . PUNCT bracis-28409 234 1 j. j. PROPN bracis-28409 234 2 develop develop PROPN bracis-28409 234 3 . . PUNCT bracis-28409 235 1 5(11 5(11 NUM bracis-28409 235 2 ) ) PUNCT bracis-28409 235 3 , , PUNCT bracis-28409 235 4 26654–26669 26654–26669 NUM bracis-28409 235 5 ( ( PUNCT bracis-28409 235 6 2019 2019 NUM bracis-28409 235 7 ) ) PUNCT bracis-28409 235 8 google google NOUN bracis-28409 235 9 scholar scholar NOUN bracis-28409 235 10     SPACE bracis-28409 235 11 vale vale NOUN bracis-28409 235 12 , , PUNCT bracis-28409 235 13 k.m.o k.m.o ADJ bracis-28409 235 14 . . PROPN bracis-28409 235 15 , , PUNCT bracis-28409 235 16 gorgônio gorgônio PROPN bracis-28409 235 17 , , PUNCT bracis-28409 235 18 a.c a.c PROPN bracis-28409 235 19 . . PROPN bracis-28409 235 20 , , PUNCT bracis-28409 235 21 da da PROPN bracis-28409 235 22 luz luz PROPN bracis-28409 235 23 , , PUNCT bracis-28409 235 24 e.g.f e.g.f PROPN bracis-28409 235 25 . . PROPN bracis-28409 235 26 , , PUNCT bracis-28409 235 27 de de PROPN bracis-28409 235 28 paula paula PROPN bracis-28409 235 29 canuto canuto PROPN bracis-28409 235 30 , , PUNCT bracis-28409 235 31 a.m. a.m. ADV bracis-28409 235 32 : : PUNCT bracis-28409 235 33 an an DET bracis-28409 235 34 efficient efficient ADJ bracis-28409 235 35 approach approach NOUN bracis-28409 235 36 to to PART bracis-28409 235 37 select select VERB bracis-28409 235 38 instances instance NOUN bracis-28409 235 39 in in ADP bracis-28409 235 40 self self NOUN bracis-28409 235 41 - - PUNCT bracis-28409 235 42 training training NOUN bracis-28409 235 43 and and CCONJ bracis-28409 235 44 co co NOUN bracis-28409 235 45 - - NOUN bracis-28409 235 46 training train VERB bracis-28409 235 47 semi semi ADJ bracis-28409 235 48 - - ADJ bracis-28409 235 49 supervised supervised ADJ bracis-28409 235 50 methods method NOUN bracis-28409 235 51 . . PUNCT bracis-28409 236 1 ieee ieee NOUN bracis-28409 236 2 access access NOUN bracis-28409 236 3 10 10 NUM bracis-28409 236 4 , , PUNCT bracis-28409 236 5 7254–7276 7254–7276 NUM bracis-28409 236 6 ( ( PUNCT bracis-28409 236 7 2021 2021 NUM bracis-28409 236 8 ) ) PUNCT bracis-28409 236 9 google google NOUN bracis-28409 236 10 scholar scholar NOUN bracis-28409 236 11     SPACE bracis-28409 236 12 gorgônio gorgônio NOUN bracis-28409 236 13 , , PUNCT bracis-28409 236 14 a.c a.c PROPN bracis-28409 236 15 . . PROPN bracis-28409 236 16 , , PUNCT bracis-28409 236 17 alves alves PROPN bracis-28409 236 18 , , PUNCT bracis-28409 236 19 c.t c.t PROPN bracis-28409 236 20 . . PROPN bracis-28409 236 21 , , PUNCT bracis-28409 236 22 lucena lucena PROPN bracis-28409 236 23 , , PUNCT bracis-28409 236 24 a.j.f a.j.f ADJ bracis-28409 236 25 . . PROPN bracis-28409 236 26 , , PUNCT bracis-28409 236 27 gorgônio gorgônio PROPN bracis-28409 236 28 , , PUNCT bracis-28409 236 29 f.l f.l PROPN bracis-28409 236 30 . . PROPN bracis-28409 236 31 , , PUNCT bracis-28409 236 32 vale vale NOUN bracis-28409 236 33 , , PUNCT bracis-28409 236 34 k.m.o k.m.o ADJ bracis-28409 236 35 . . PROPN bracis-28409 236 36 , , PUNCT bracis-28409 236 37 canuto canuto PROPN bracis-28409 236 38 , , PUNCT bracis-28409 236 39 a.m.p a.m.p PROPN bracis-28409 236 40 . . PUNCT bracis-28409 236 41 : : PUNCT bracis-28409 237 1 analysis analysis NOUN bracis-28409 237 2 of of ADP bracis-28409 237 3 the the DET bracis-28409 237 4 threshold threshold NOUN bracis-28409 237 5 variation variation NOUN bracis-28409 237 6 of of ADP bracis-28409 237 7 the the DET bracis-28409 237 8 flexcon flexcon NOUN bracis-28409 237 9 - - PUNCT bracis-28409 237 10 c c NOUN bracis-28409 237 11 algorithm algorithm NOUN bracis-28409 237 12 for for ADP bracis-28409 237 13 semi semi ADJ bracis-28409 237 14 - - ADJ bracis-28409 237 15 supervised supervised ADJ bracis-28409 237 16 learning learning NOUN bracis-28409 237 17 . . PUNCT bracis-28409 238 1 in in ADP bracis-28409 238 2 : : PUNCT bracis-28409 238 3 anais anais PROPN bracis-28409 238 4 do do AUX bracis-28409 238 5 xv xv PROPN bracis-28409 238 6 encontro encontro VERB bracis-28409 238 7 nacional nacional ADJ bracis-28409 238 8 de de X bracis-28409 238 9 inteligência inteligência PROPN bracis-28409 238 10 artificial artificial PROPN bracis-28409 238 11 e e NOUN bracis-28409 238 12 computacional computacional NOUN bracis-28409 238 13 , , PUNCT bracis-28409 238 14 pp pp ADJ bracis-28409 238 15 . . PUNCT bracis-28409 239 1 775–786 775–786 NUM bracis-28409 239 2 . . PUNCT bracis-28409 240 1 sbc sbc NOUN bracis-28409 240 2 ( ( PUNCT bracis-28409 240 3 2018 2018 NUM bracis-28409 240 4 ) ) PUNCT bracis-28409 240 5 google google NOUN bracis-28409 240 6 scholar scholar NOUN bracis-28409 240 7     SPACE bracis-28409 240 8 vale vale NOUN bracis-28409 240 9 , , PUNCT bracis-28409 240 10 k.m.o k.m.o NOUN bracis-28409 240 11 . . PROPN bracis-28409 240 12 , , PUNCT bracis-28409 240 13 et et PROPN bracis-28409 240 14 al al PROPN bracis-28409 240 15 . . PUNCT bracis-28409 240 16 : : PUNCT bracis-28409 241 1 a a DET bracis-28409 241 2 data data NOUN bracis-28409 241 3 stratification stratification NOUN bracis-28409 241 4 process process NOUN bracis-28409 241 5 for for ADP bracis-28409 241 6 instances instance NOUN bracis-28409 241 7 selection selection NOUN bracis-28409 241 8 in in ADP bracis-28409 241 9 semi semi ADJ bracis-28409 241 10 - - ADJ bracis-28409 241 11 supervised supervised ADJ bracis-28409 241 12 learning learning NOUN bracis-28409 241 13 . . PUNCT bracis-28409 242 1 in in ADP bracis-28409 242 2 : : PUNCT bracis-28409 242 3 2019 2019 NUM bracis-28409 242 4 international international ADJ bracis-28409 242 5 joint joint ADJ bracis-28409 242 6 conference conference NOUN bracis-28409 242 7 on on ADP bracis-28409 242 8 neural neural ADJ bracis-28409 242 9 networks network NOUN bracis-28409 242 10 ( ( PUNCT bracis-28409 242 11 ijcnn ijcnn PROPN bracis-28409 242 12 ) ) PUNCT bracis-28409 242 13 , , PUNCT bracis-28409 242 14 pp pp ADP bracis-28409 242 15 . . PUNCT bracis-28409 243 1 1–8 1–8 X bracis-28409 243 2 . . PUNCT bracis-28409 243 3 ieee ieee PROPN bracis-28409 243 4 ( ( PUNCT bracis-28409 243 5 2019 2019 NUM bracis-28409 243 6 ) ) PUNCT bracis-28409 243 7 google google NOUN bracis-28409 243 8 scholar scholar NOUN bracis-28409 243 9     SPACE bracis-28409 243 10 breiman breiman NOUN bracis-28409 243 11 , , PUNCT bracis-28409 243 12 l. l. PROPN bracis-28409 243 13 : : PUNCT bracis-28409 243 14 bias bias PROPN bracis-28409 243 15 , , PUNCT bracis-28409 243 16 variance variance NOUN bracis-28409 243 17 , , PUNCT bracis-28409 243 18 and and CCONJ bracis-28409 243 19 arcing arc VERB bracis-28409 243 20 classifiers classifier NOUN bracis-28409 243 21 . . PUNCT bracis-28409 244 1 technical technical ADJ bracis-28409 244 2 report report NOUN bracis-28409 244 3 , , PUNCT bracis-28409 244 4 tech tech PROPN bracis-28409 244 5 . . PUNCT bracis-28409 245 1 rep rep PROPN bracis-28409 245 2 . . PROPN bracis-28409 245 3 460 460 NUM bracis-28409 245 4 , , PUNCT bracis-28409 245 5 statistics statistics PROPN bracis-28409 245 6 department department PROPN bracis-28409 245 7 , , PUNCT bracis-28409 245 8 university university PROPN bracis-28409 245 9 of of ADP bracis-28409 245 10 california california PROPN bracis-28409 245 11 , , PUNCT bracis-28409 245 12 berkeley berkeley PROPN bracis-28409 245 13 ( ( PUNCT bracis-28409 245 14 1996 1996 NUM bracis-28409 245 15 ) ) PUNCT bracis-28409 245 16 google google PROPN bracis-28409 245 17 scholar scholar NOUN bracis-28409 245 18     SPACE bracis-28409 245 19 dheeru dheeru PROPN bracis-28409 245 20 , , PUNCT bracis-28409 245 21 d. d. PROPN bracis-28409 245 22 , , PUNCT bracis-28409 245 23 karra karra PROPN bracis-28409 245 24 taniskidou taniskidou NOUN bracis-28409 245 25 , , PUNCT bracis-28409 245 26 e. e. PROPN bracis-28409 245 27 : : PUNCT bracis-28409 245 28 uci uci PROPN bracis-28409 245 29 machine machine NOUN bracis-28409 245 30 learning learn VERB bracis-28409 245 31 repository repository NOUN bracis-28409 245 32 ( ( PUNCT bracis-28409 245 33 2017 2017 NUM bracis-28409 245 34 ) ) PUNCT bracis-28409 245 35 google google PROPN bracis-28409 245 36 scholar scholar NOUN bracis-28409 245 37     SPACE bracis-28409 245 38 smith smith PROPN bracis-28409 245 39 , , PUNCT bracis-28409 245 40 j.w j.w PROPN bracis-28409 245 41 . . PROPN bracis-28409 245 42 , , PUNCT bracis-28409 245 43 everhart everhart PROPN bracis-28409 245 44 , , PUNCT bracis-28409 245 45 j.e j.e PROPN bracis-28409 245 46 . . PROPN bracis-28409 245 47 , , PUNCT bracis-28409 245 48 dickson dickson PROPN bracis-28409 245 49 , , PUNCT bracis-28409 245 50 w.c w.c PROPN bracis-28409 245 51 . . PROPN bracis-28409 245 52 , , PUNCT bracis-28409 245 53 knowler knowler PROPN bracis-28409 245 54 , , PUNCT bracis-28409 245 55 w.c w.c PROPN bracis-28409 245 56 . . PROPN bracis-28409 245 57 , , PUNCT bracis-28409 245 58 johannes johannes PROPN bracis-28409 245 59 , , PUNCT bracis-28409 245 60 r.s r.s PROPN bracis-28409 245 61 . . PROPN bracis-28409 245 62 : : PUNCT bracis-28409 246 1 using use VERB bracis-28409 246 2 the the DET bracis-28409 246 3 adap adap NOUN bracis-28409 246 4 learning learn VERB bracis-28409 246 5 algorithm algorithm NOUN bracis-28409 246 6 to to PART bracis-28409 246 7 forecast forecast VERB bracis-28409 246 8 the the DET bracis-28409 246 9 onset onset NOUN bracis-28409 246 10 of of ADP bracis-28409 246 11 diabetes diabetes NOUN bracis-28409 246 12 mellitus mellitus NOUN bracis-28409 246 13 . . PUNCT bracis-28409 247 1 in in ADP bracis-28409 247 2 : : PUNCT bracis-28409 247 3 proceedings proceeding NOUN bracis-28409 247 4 of of ADP bracis-28409 247 5 the the DET bracis-28409 247 6 annual annual ADJ bracis-28409 247 7 symposium symposium NOUN bracis-28409 247 8 on on ADP bracis-28409 247 9 computer computer NOUN bracis-28409 247 10 application application NOUN bracis-28409 247 11 in in ADP bracis-28409 247 12 medical medical ADJ bracis-28409 247 13 care care NOUN bracis-28409 247 14 , , PUNCT bracis-28409 247 15 p. p. NOUN bracis-28409 247 16 261 261 NUM bracis-28409 247 17 . . PUNCT bracis-28409 248 1 american american PROPN bracis-28409 248 2 medical medical PROPN bracis-28409 248 3 informatics informatics PROPN bracis-28409 248 4 association association PROPN bracis-28409 248 5 ( ( PUNCT bracis-28409 248 6 1988 1988 NUM bracis-28409 248 7 ) ) PUNCT bracis-28409 248 8 google google NOUN bracis-28409 248 9 scholar scholar NOUN bracis-28409 248 10     SPACE bracis-28409 248 11 pölsterl pölsterl NOUN bracis-28409 248 12 , , PUNCT bracis-28409 248 13 s. s. PROPN bracis-28409 248 14 : : PUNCT bracis-28409 248 15 scikit scikit NOUN bracis-28409 248 16 - - PUNCT bracis-28409 248 17 survival survival PROPN bracis-28409 248 18 : : PUNCT bracis-28409 248 19 a a DET bracis-28409 248 20 library library NOUN bracis-28409 248 21 for for ADP bracis-28409 248 22 time time NOUN bracis-28409 248 23 - - PUNCT bracis-28409 248 24 to to ADP bracis-28409 248 25 - - PUNCT bracis-28409 248 26 event event NOUN bracis-28409 248 27 analysis analysis NOUN bracis-28409 248 28 built build VERB bracis-28409 248 29 on on ADP bracis-28409 248 30 top top NOUN bracis-28409 248 31 of of ADP bracis-28409 248 32 scikit scikit NOUN bracis-28409 248 33 - - PUNCT bracis-28409 248 34 learn learn NOUN bracis-28409 248 35 . . PUNCT bracis-28409 249 1 j. j. PROPN bracis-28409 249 2 mach mach PROPN bracis-28409 249 3 . . PUNCT bracis-28409 250 1 learn learn VERB bracis-28409 250 2 . . PUNCT bracis-28409 251 1 res re NOUN bracis-28409 251 2 . . PUNCT bracis-28409 252 1 21(1 21(1 NUM bracis-28409 252 2 ) ) PUNCT bracis-28409 252 3 , , PUNCT bracis-28409 252 4 8747–8752 8747–8752 NUM bracis-28409 252 5 ( ( PUNCT bracis-28409 252 6 2020 2020 NUM bracis-28409 252 7 ) ) PUNCT bracis-28409 252 8 google google PROPN bracis-28409 252 9 scholar scholar NOUN bracis-28409 252 10     SPACE bracis-28409 252 11 bisong bisong NOUN bracis-28409 252 12 , , PUNCT bracis-28409 252 13 e. e. PROPN bracis-28409 252 14 : : PUNCT bracis-28409 252 15 building build VERB bracis-28409 252 16 machine machine NOUN bracis-28409 252 17 learning learning NOUN bracis-28409 252 18 and and CCONJ bracis-28409 252 19 deep deep ADJ bracis-28409 252 20 learning learning NOUN bracis-28409 252 21 models model NOUN bracis-28409 252 22 on on ADP bracis-28409 252 23 google google PROPN bracis-28409 252 24 cloud cloud PROPN bracis-28409 252 25 platform platform NOUN bracis-28409 252 26 . . PUNCT bracis-28409 253 1 apress apress PROPN bracis-28409 253 2 , , PUNCT bracis-28409 253 3 berkeley berkeley PROPN bracis-28409 253 4 ( ( PUNCT bracis-28409 253 5 2019 2019 NUM bracis-28409 253 6 ) ) PUNCT bracis-28409 253 7 . . PUNCT bracis-28409 254 1 https://doi.org/10.1007/978-1-4842-4470-8 https://doi.org/10.1007/978-1-4842-4470-8 PROPN bracis-28409 254 2 harris harris PROPN bracis-28409 254 3 , , PUNCT bracis-28409 254 4 c.r c.r PROPN bracis-28409 254 5 . . PROPN bracis-28409 254 6 , , PUNCT bracis-28409 254 7 et et PROPN bracis-28409 254 8 al al PROPN bracis-28409 254 9 . . PUNCT bracis-28409 254 10 : : PUNCT bracis-28409 254 11 array array NOUN bracis-28409 254 12 programming program VERB bracis-28409 254 13 with with ADP bracis-28409 254 14 numpy numpy NOUN bracis-28409 254 15 . . PUNCT bracis-28409 254 16 nature nature NOUN bracis-28409 254 17 585(7825 585(7825 PROPN bracis-28409 254 18 ) ) PUNCT bracis-28409 254 19 , , PUNCT bracis-28409 254 20 357–362 357–362 NUM bracis-28409 254 21 ( ( PUNCT bracis-28409 254 22 2020 2020 NUM bracis-28409 254 23 ) ) PUNCT bracis-28409 254 24 google google PROPN bracis-28409 254 25 scholar scholar NOUN bracis-28409 254 26     SPACE bracis-28409 254 27 nelli nelli PROPN bracis-28409 254 28 , , PUNCT bracis-28409 254 29 f. f. PROPN bracis-28409 254 30 : : PUNCT bracis-28409 254 31 python python PROPN bracis-28409 254 32 data datum NOUN bracis-28409 254 33 analytics analytic NOUN bracis-28409 254 34 with with ADP bracis-28409 254 35 pandas panda NOUN bracis-28409 254 36 , , PUNCT bracis-28409 254 37 numpy numpy NOUN bracis-28409 254 38 , , PUNCT bracis-28409 254 39 and and CCONJ bracis-28409 254 40 matplotlib matplotlib PROPN bracis-28409 254 41 ( ( PUNCT bracis-28409 254 42 2018 2018 NUM bracis-28409 254 43 ) ) PUNCT bracis-28409 254 44 google google PROPN bracis-28409 254 45 scholar scholar NOUN bracis-28409 254 46     SPACE bracis-28409 254 47 araújo araújo PROPN bracis-28409 254 48 , , PUNCT bracis-28409 254 49 y.n y.n PROPN bracis-28409 254 50 . . PROPN bracis-28409 254 51 , , PUNCT bracis-28409 254 52 et et PROPN bracis-28409 254 53 al al PROPN bracis-28409 254 54 . . PUNCT bracis-28409 254 55 : : PUNCT bracis-28409 255 1 a a DET bracis-28409 255 2 data data NOUN bracis-28409 255 3 stratification stratification NOUN bracis-28409 255 4 process process NOUN bracis-28409 255 5 for for ADP bracis-28409 255 6 instances instance NOUN bracis-28409 255 7 selection selection NOUN bracis-28409 255 8 applied apply VERB bracis-28409 255 9 to to ADP bracis-28409 255 10 co co VERB bracis-28409 255 11 - - NOUN bracis-28409 255 12 training train VERB bracis-28409 255 13 semi semi ADJ bracis-28409 255 14 - - ADJ bracis-28409 255 15 supervised supervised ADJ bracis-28409 255 16 learning learning NOUN bracis-28409 255 17 algorithm algorithm NOUN bracis-28409 255 18 . . PUNCT bracis-28409 256 1 in in ADP bracis-28409 256 2 : : PUNCT bracis-28409 256 3 2021 2021 NUM bracis-28409 256 4 international international ADJ bracis-28409 256 5 joint joint ADJ bracis-28409 256 6 conference conference NOUN bracis-28409 256 7 on on ADP bracis-28409 256 8 neural neural ADJ bracis-28409 256 9 networks network NOUN bracis-28409 256 10 ( ( PUNCT bracis-28409 256 11 ijcnn ijcnn PROPN bracis-28409 256 12 ) ) PUNCT bracis-28409 256 13 , , PUNCT bracis-28409 256 14 pp pp ADP bracis-28409 256 15 . . PUNCT bracis-28409 257 1 1–8 1–8 X bracis-28409 257 2 . . PUNCT bracis-28409 257 3 ieee ieee NOUN bracis-28409 257 4 ( ( PUNCT bracis-28409 257 5 2021 2021 NUM bracis-28409 257 6 ) ) PUNCT bracis-28409 257 7 google google NOUN bracis-28409 257 8 scholar scholar NOUN bracis-28409 257 9     SPACE bracis-28409 257 10 vale vale NOUN bracis-28409 257 11 , , PUNCT bracis-28409 257 12 k.m.o k.m.o ADJ bracis-28409 257 13 . . PROPN bracis-28409 257 14 , , PUNCT bracis-28409 257 15 gorgônio gorgônio PROPN bracis-28409 257 16 , , PUNCT bracis-28409 257 17 f.l f.l PROPN bracis-28409 257 18 . . PROPN bracis-28409 257 19 , , PUNCT bracis-28409 257 20 araújo araújo PROPN bracis-28409 257 21 , , PUNCT bracis-28409 257 22 y.n y.n PROPN bracis-28409 257 23 . . PROPN bracis-28409 257 24 , , PUNCT bracis-28409 257 25 gorgônio gorgônio PROPN bracis-28409 257 26 , , PUNCT bracis-28409 257 27 a.c a.c PROPN bracis-28409 257 28 . . PROPN bracis-28409 257 29 , , PUNCT bracis-28409 257 30 de de PROPN bracis-28409 257 31 p p PROPN bracis-28409 257 32 canuto canuto NOUN bracis-28409 257 33 , , PUNCT bracis-28409 257 34 a.m. a.m. ADV bracis-28409 257 35 : : PUNCT bracis-28409 257 36 a a DET bracis-28409 257 37 co co ADJ bracis-28409 257 38 - - ADJ bracis-28409 257 39 training training NOUN bracis-28409 257 40 - - PUNCT bracis-28409 257 41 based base VERB bracis-28409 257 42 algorithm algorithm NOUN bracis-28409 257 43 using use VERB bracis-28409 257 44 confidence confidence NOUN bracis-28409 257 45 values value NOUN bracis-28409 257 46 to to PART bracis-28409 257 47 select select VERB bracis-28409 257 48 instances instance NOUN bracis-28409 257 49 . . PUNCT bracis-28409 258 1 in in ADP bracis-28409 258 2 : : PUNCT bracis-28409 258 3 2020 2020 NUM bracis-28409 258 4 international international ADJ bracis-28409 258 5 joint joint ADJ bracis-28409 258 6 conference conference NOUN bracis-28409 258 7 on on ADP bracis-28409 258 8 neural neural ADJ bracis-28409 258 9 networks network NOUN bracis-28409 258 10 ( ( PUNCT bracis-28409 258 11 ijcnn ijcnn PROPN bracis-28409 258 12 ) ) PUNCT bracis-28409 258 13 , , PUNCT bracis-28409 258 14 pp pp ADP bracis-28409 258 15 . . PUNCT bracis-28409 259 1 1–7 1–7 X bracis-28409 259 2 . . PUNCT bracis-28409 259 3 ieee ieee NOUN bracis-28409 259 4 ( ( PUNCT bracis-28409 259 5 2020 2020 NUM bracis-28409 259 6 ) ) PUNCT bracis-28409 259 7 google google PROPN bracis-28409 259 8 scholar scholar NOUN bracis-28409 259 9     SPACE bracis-28409 259 10 kuncheva kuncheva PROPN bracis-28409 259 11 , , PUNCT bracis-28409 259 12 l.i l.i PROPN bracis-28409 259 13 . . PROPN bracis-28409 259 14 : : PUNCT bracis-28409 259 15 combining combine VERB bracis-28409 259 16 pattern pattern NOUN bracis-28409 259 17 classifiers classifier NOUN bracis-28409 259 18 : : PUNCT bracis-28409 259 19 methods method NOUN bracis-28409 259 20 and and CCONJ bracis-28409 259 21 algorithms algorithm NOUN bracis-28409 259 22 . . PUNCT bracis-28409 260 1 john john PROPN bracis-28409 260 2 wiley wiley PROPN bracis-28409 260 3 & & CCONJ bracis-28409 260 4 sons sons PROPN bracis-28409 260 5 ( ( PUNCT bracis-28409 260 6 2014 2014 NUM bracis-28409 260 7 ) ) PUNCT bracis-28409 260 8 google google PROPN bracis-28409 260 9 scholar scholar NOUN bracis-28409 260 10     SPACE bracis-28409 260 11 theodorsson theodorsson NOUN bracis-28409 260 12 - - PUNCT bracis-28409 260 13 norheim norheim PROPN bracis-28409 260 14 , , PUNCT bracis-28409 260 15 e. e. PROPN bracis-28409 260 16 : : PUNCT bracis-28409 260 17 friedman friedman PROPN bracis-28409 260 18 and and CCONJ bracis-28409 260 19 quade quade PROPN bracis-28409 260 20 tests test NOUN bracis-28409 260 21 : : PUNCT bracis-28409 260 22 basic basic ADJ bracis-28409 260 23 computer computer NOUN bracis-28409 260 24 program program NOUN bracis-28409 260 25 to to PART bracis-28409 260 26 perform perform VERB bracis-28409 260 27 nonparametric nonparametric NOUN bracis-28409 260 28 two two NUM bracis-28409 260 29 - - PUNCT bracis-28409 260 30 way way NOUN bracis-28409 260 31 analysis analysis NOUN bracis-28409 260 32 of of ADP bracis-28409 260 33 variance variance NOUN bracis-28409 260 34 and and CCONJ bracis-28409 260 35 multiple multiple ADJ bracis-28409 260 36 comparisons comparison NOUN bracis-28409 260 37 on on ADP bracis-28409 260 38 ranks rank NOUN bracis-28409 260 39 of of ADP bracis-28409 260 40 several several ADJ bracis-28409 260 41 related related ADJ bracis-28409 260 42 samples sample NOUN bracis-28409 260 43 . . PUNCT bracis-28409 261 1 comput comput NOUN bracis-28409 261 2 . . PUNCT bracis-28409 262 1 biol biol PROPN bracis-28409 262 2 . . PUNCT bracis-28409 263 1 med med PROPN bracis-28409 263 2 . . PUNCT bracis-28409 264 1 17(2 17(2 NUM bracis-28409 264 2 ) ) PUNCT bracis-28409 264 3 , , PUNCT bracis-28409 264 4 85–99 85–99 NUM bracis-28409 264 5 ( ( PUNCT bracis-28409 264 6 1987 1987 NUM bracis-28409 264 7 ) ) PUNCT bracis-28409 264 8 article article NOUN bracis-28409 264 9     SPACE bracis-28409 264 10 google google PROPN bracis-28409 264 11 scholar scholar NOUN bracis-28409 264 12     SPACE bracis-28409 264 13 download download NOUN bracis-28409 264 14 references reference NOUN bracis-28409 264 15 author author NOUN bracis-28409 264 16 information information NOUN bracis-28409 264 17 authors author NOUN bracis-28409 264 18 and and CCONJ bracis-28409 264 19 affiliations affiliation NOUN bracis-28409 264 20 department department PROPN bracis-28409 264 21 of of ADP bracis-28409 264 22 computing computing NOUN bracis-28409 264 23 and and CCONJ bracis-28409 264 24 technology technology NOUN bracis-28409 264 25 ( ( PUNCT bracis-28409 264 26 dct dct PROPN bracis-28409 264 27 ) ) PUNCT bracis-28409 264 28 , , PUNCT bracis-28409 264 29 federal federal ADJ bracis-28409 264 30 university university PROPN bracis-28409 264 31 of of ADP bracis-28409 264 32 rio rio PROPN bracis-28409 264 33 grande grande PROPN bracis-28409 264 34 do do AUX bracis-28409 264 35 norte norte NOUN bracis-28409 264 36 ( ( PUNCT bracis-28409 264 37 ufrn ufrn PROPN bracis-28409 264 38 ) ) PUNCT bracis-28409 264 39 , , PUNCT bracis-28409 264 40 caicó caicó PROPN bracis-28409 264 41 , , PUNCT bracis-28409 264 42 rn rn PROPN bracis-28409 264 43 , , PUNCT bracis-28409 264 44 brazil brazil PROPN bracis-28409 264 45 arthur arthur PROPN bracis-28409 264 46 medeiros medeiros PROPN bracis-28409 264 47 , , PUNCT bracis-28409 264 48     SPACE bracis-28409 264 49 karliane karliane PROPN bracis-28409 264 50 medeiros medeiros PROPN bracis-28409 264 51 ovidio ovidio PROPN bracis-28409 264 52 vale vale NOUN bracis-28409 264 53 , , PUNCT bracis-28409 264 54     SPACE bracis-28409 264 55 flavius flavius PROPN bracis-28409 264 56 l. l. PROPN bracis-28409 264 57 gorgônio gorgônio PROPN bracis-28409 264 58     SPACE bracis-28409 264 59 & & CCONJ bracis-28409 264 60     SPACE bracis-28409 264 61 anne anne PROPN bracis-28409 264 62 magály magály PROPN bracis-28409 264 63 de de PROPN bracis-28409 264 64 paula paula PROPN bracis-28409 264 65 canuto canuto PROPN bracis-28409 264 66 department department PROPN bracis-28409 264 67 of of ADP bracis-28409 264 68 informatics informatic NOUN bracis-28409 264 69 and and CCONJ bracis-28409 264 70 applied apply VERB bracis-28409 264 71 mathematics mathematic NOUN bracis-28409 264 72 ( ( PUNCT bracis-28409 264 73 dimap dimap NOUN bracis-28409 264 74 ) ) PUNCT bracis-28409 264 75 , , PUNCT bracis-28409 264 76 federal federal ADJ bracis-28409 264 77 university university PROPN bracis-28409 264 78 of of ADP bracis-28409 264 79 rio rio PROPN bracis-28409 264 80 grande grande PROPN bracis-28409 264 81 do do VERB bracis-28409 264 82 norte norte NOUN bracis-28409 264 83 ( ( PUNCT bracis-28409 264 84 ufrn ufrn PROPN bracis-28409 264 85 ) ) PUNCT bracis-28409 264 86 , , PUNCT bracis-28409 264 87 natal natal NOUN bracis-28409 264 88 , , PUNCT bracis-28409 264 89 rn rn PROPN bracis-28409 264 90 , , PUNCT bracis-28409 264 91 brazil brazil PROPN bracis-28409 264 92 arthur arthur PROPN bracis-28409 264 93 c. c. PROPN bracis-28409 264 94 gorgônio gorgônio PROPN bracis-28409 264 95 authors author NOUN bracis-28409 264 96 arthur arthur PROPN bracis-28409 264 97 medeirosview medeirosview PROPN bracis-28409 264 98 author author NOUN bracis-28409 264 99 publications publication NOUN bracis-28409 264 100 search search NOUN bracis-28409 264 101 author author NOUN bracis-28409 264 102 on on ADP bracis-28409 264 103 : : PUNCT bracis-28409 264 104 pubmed pubmed PROPN bracis-28409 264 105     SPACE bracis-28409 264 106 google google PROPN bracis-28409 264 107 scholar scholar PROPN bracis-28409 264 108 arthur arthur PROPN bracis-28409 264 109 c. c. PROPN bracis-28409 264 110 gorgônioview gorgônioview PROPN bracis-28409 264 111 author author NOUN bracis-28409 264 112 publications publication NOUN bracis-28409 264 113 search search NOUN bracis-28409 264 114 author author NOUN bracis-28409 264 115 on on ADP bracis-28409 264 116 : : PUNCT bracis-28409 264 117 pubmed pubmed PROPN bracis-28409 264 118     SPACE bracis-28409 264 119 google google PROPN bracis-28409 264 120 scholar scholar PROPN bracis-28409 264 121 karliane karliane PROPN bracis-28409 264 122 medeiros medeiros PROPN bracis-28409 264 123 ovidio ovidio ADJ bracis-28409 264 124 valeview valeview NOUN bracis-28409 264 125 author author NOUN bracis-28409 264 126 publications publication NOUN bracis-28409 264 127 search search NOUN bracis-28409 264 128 author author NOUN bracis-28409 264 129 on on ADP bracis-28409 264 130 : : PUNCT bracis-28409 264 131 pubmed pubmed PROPN bracis-28409 264 132     SPACE bracis-28409 264 133 google google PROPN bracis-28409 264 134 scholar scholar NOUN bracis-28409 264 135 flavius flavius PROPN bracis-28409 264 136 l. l. PROPN bracis-28409 264 137 gorgônioview gorgônioview PROPN bracis-28409 264 138 author author NOUN bracis-28409 264 139 publications publication NOUN bracis-28409 264 140 search search NOUN bracis-28409 264 141 author author NOUN bracis-28409 264 142 on on ADP bracis-28409 264 143 : : PUNCT bracis-28409 264 144 pubmed pubmed PROPN bracis-28409 264 145     SPACE bracis-28409 264 146 google google PROPN bracis-28409 264 147 scholar scholar PROPN bracis-28409 264 148 anne anne PROPN bracis-28409 264 149 magály magály PROPN bracis-28409 264 150 de de PROPN bracis-28409 264 151 paula paula PROPN bracis-28409 264 152 canutoview canutoview PROPN bracis-28409 264 153 author author NOUN bracis-28409 264 154 publications publication NOUN bracis-28409 264 155 search search NOUN bracis-28409 264 156 author author NOUN bracis-28409 264 157 on on ADP bracis-28409 264 158 : : PUNCT bracis-28409 264 159 pubmed pubmed PROPN bracis-28409 264 160     SPACE bracis-28409 264 161 google google PROPN bracis-28409 264 162 scholar scholar NOUN bracis-28409 264 163 corresponding correspond VERB bracis-28409 264 164 author author NOUN bracis-28409 264 165 correspondence correspondence NOUN bracis-28409 264 166 to to ADP bracis-28409 264 167 arthur arthur PROPN bracis-28409 264 168 medeiros medeiros PROPN bracis-28409 264 169 . . PUNCT bracis-28409 265 1 editor editor NOUN bracis-28409 265 2 information information NOUN bracis-28409 265 3 editors editor NOUN bracis-28409 265 4 and and CCONJ bracis-28409 265 5 affiliations affiliation NOUN bracis-28409 265 6 federal federal PROPN bracis-28409 265 7 university university PROPN bracis-28409 265 8 of of ADP bracis-28409 265 9 são são PROPN bracis-28409 265 10 carlos carlos PROPN bracis-28409 265 11 , , PUNCT bracis-28409 265 12 são são PROPN bracis-28409 265 13 carlos carlos PROPN bracis-28409 265 14 , , PUNCT bracis-28409 265 15 brazil brazil PROPN bracis-28409 265 16 murilo murilo PROPN bracis-28409 265 17 c. c. PROPN bracis-28409 265 18 naldi naldi PROPN bracis-28409 265 19 centro centro PROPN bracis-28409 265 20 universitario universitario PROPN bracis-28409 265 21 da da PROPN bracis-28409 265 22 fei fei PROPN bracis-28409 265 23 , , PUNCT bracis-28409 265 24 são são PROPN bracis-28409 265 25 bernardo bernardo PROPN bracis-28409 265 26 do do AUX bracis-28409 265 27 campo campo PROPN bracis-28409 265 28 , , PUNCT bracis-28409 265 29 brazil brazil PROPN bracis-28409 265 30 reinaldo reinaldo PROPN bracis-28409 265 31 a. a. PROPN bracis-28409 265 32 c. c. PROPN bracis-28409 265 33 bianchi bianchi PROPN bracis-28409 265 34 rights right NOUN bracis-28409 265 35 and and CCONJ bracis-28409 265 36 permissions permission NOUN bracis-28409 265 37 reprints reprint NOUN bracis-28409 265 38 and and CCONJ bracis-28409 265 39 permissions permission VERB bracis-28409 265 40 copyright copyright NOUN bracis-28409 265 41 information information NOUN bracis-28409 265 42 © © ADP bracis-28409 265 43 2023 2023 NUM bracis-28409 265 44 the the DET bracis-28409 265 45 author(s author(s NOUN bracis-28409 265 46 ) ) PUNCT bracis-28409 265 47 , , PUNCT bracis-28409 265 48 under under ADP bracis-28409 265 49 exclusive exclusive ADJ bracis-28409 265 50 license license NOUN bracis-28409 265 51 to to ADP bracis-28409 265 52 springer springer NOUN bracis-28409 265 53 nature nature PROPN bracis-28409 265 54 switzerland switzerland PROPN bracis-28409 265 55 ag ag PROPN bracis-28409 265 56 about about ADP bracis-28409 265 57 this this DET bracis-28409 265 58 paper paper NOUN bracis-28409 265 59 cite cite VERB bracis-28409 265 60 this this DET bracis-28409 265 61 paper paper NOUN bracis-28409 265 62 medeiros medeiros PROPN bracis-28409 265 63 , , PUNCT bracis-28409 265 64 a. a. PROPN bracis-28409 265 65 , , PUNCT bracis-28409 265 66 gorgônio gorgônio PROPN bracis-28409 265 67 , , PUNCT bracis-28409 265 68 a.c a.c PROPN bracis-28409 265 69 . . PROPN bracis-28409 265 70 , , PUNCT bracis-28409 265 71 vale vale NOUN bracis-28409 265 72 , , PUNCT bracis-28409 265 73 k.m.o k.m.o ADJ bracis-28409 265 74 . . PROPN bracis-28409 265 75 , , PUNCT bracis-28409 265 76 gorgônio gorgônio PROPN bracis-28409 265 77 , , PUNCT bracis-28409 265 78 f.l f.l PROPN bracis-28409 265 79 . . PROPN bracis-28409 265 80 , , PUNCT bracis-28409 265 81 canuto canuto PROPN bracis-28409 265 82 , , PUNCT bracis-28409 265 83 a.m.d.p a.m.d.p PROPN bracis-28409 265 84 . . PROPN bracis-28409 265 85 ( ( PUNCT bracis-28409 265 86 2023 2023 NUM bracis-28409 265 87 ) ) PUNCT bracis-28409 265 88 . . PUNCT bracis-28409 266 1 flexcon flexcon NOUN bracis-28409 266 2 - - PUNCT bracis-28409 266 3 ce ce PROPN bracis-28409 266 4 : : PUNCT bracis-28409 266 5 a a DET bracis-28409 266 6 semi semi ADJ bracis-28409 266 7 - - ADJ bracis-28409 266 8 supervised supervised ADJ bracis-28409 266 9 method method NOUN bracis-28409 266 10 with with ADP bracis-28409 266 11     SPACE bracis-28409 266 12 an an DET bracis-28409 266 13     SPACE bracis-28409 266 14 ensemble ensemble ADJ bracis-28409 266 15 - - PUNCT bracis-28409 266 16 based base VERB bracis-28409 266 17 adaptive adaptive ADJ bracis-28409 266 18 confidence confidence NOUN bracis-28409 266 19 . . PUNCT bracis-28409 267 1 in in ADP bracis-28409 267 2 : : PUNCT bracis-28409 267 3 naldi naldi PROPN bracis-28409 267 4 , , PUNCT bracis-28409 267 5 m.c m.c PROPN bracis-28409 267 6 . . PROPN bracis-28409 267 7 , , PUNCT bracis-28409 267 8 bianchi bianchi PROPN bracis-28409 267 9 , , PUNCT bracis-28409 267 10 r.a.c r.a.c ADP bracis-28409 267 11 . . PUNCT bracis-28409 267 12 ( ( PUNCT bracis-28409 267 13 eds ed NOUN bracis-28409 267 14 ) ) PUNCT bracis-28409 267 15 intelligent intelligent ADJ bracis-28409 267 16 systems system NOUN bracis-28409 267 17 . . PUNCT bracis-28409 268 1 bracis bracis PROPN bracis-28409 268 2 2023 2023 NUM bracis-28409 268 3 . . PUNCT bracis-28409 269 1 lecture lecture NOUN bracis-28409 269 2 notes note NOUN bracis-28409 269 3 in in ADP bracis-28409 269 4 computer computer NOUN bracis-28409 269 5 science science NOUN bracis-28409 269 6 ( ( PUNCT bracis-28409 269 7 ) ) PUNCT bracis-28409 269 8 , , PUNCT bracis-28409 269 9 vol vol NOUN bracis-28409 269 10 14197 14197 NUM bracis-28409 269 11 . . PUNCT bracis-28409 270 1 springer springer NOUN bracis-28409 270 2 , , PUNCT bracis-28409 270 3 cham cham PROPN bracis-28409 270 4 . . PUNCT bracis-28409 271 1 https://doi.org/10.1007/978-3-031-45392-2_7 https://doi.org/10.1007/978-3-031-45392-2_7 NOUN bracis-28409 271 2 download download NOUN bracis-28409 271 3 citation citation NOUN bracis-28409 271 4 .ris .ris PUNCT bracis-28409 272 1 .enw .enw PROPN bracis-28409 272 2 .bib .bib PUNCT bracis-28409 273 1 doi doi PROPN bracis-28409 273 2 : : PUNCT bracis-28409 273 3 https://doi.org/10.1007/978-3-031-45392-2_7 https://doi.org/10.1007/978-3-031-45392-2_7 NOUN bracis-28409 273 4 published publish VERB bracis-28409 273 5 : : PUNCT bracis-28409 273 6 12 12 NUM bracis-28409 273 7 october october PROPN bracis-28409 273 8 2023 2023 NUM bracis-28409 273 9 publisher publisher NOUN bracis-28409 273 10 name name NOUN bracis-28409 273 11 : : PUNCT bracis-28409 273 12 springer springer NOUN bracis-28409 273 13 , , PUNCT bracis-28409 273 14 cham cham PROPN bracis-28409 273 15 print print PROPN bracis-28409 273 16 isbn isbn PROPN bracis-28409 273 17 : : PUNCT bracis-28409 273 18 978 978 NUM bracis-28409 273 19 - - SYM bracis-28409 273 20 3 3 NUM bracis-28409 273 21 - - PUNCT bracis-28409 273 22 031 031 NUM bracis-28409 273 23 - - PUNCT bracis-28409 273 24 45391 45391 NUM bracis-28409 273 25 - - SYM bracis-28409 273 26 5 5 NUM bracis-28409 273 27 online online ADJ bracis-28409 273 28 isbn isbn NOUN bracis-28409 273 29 : : PUNCT bracis-28409 273 30 978 978 NUM bracis-28409 273 31 - - SYM bracis-28409 273 32 3 3 NUM bracis-28409 273 33 - - PUNCT bracis-28409 273 34 031 031 NUM bracis-28409 273 35 - - PUNCT bracis-28409 273 36 45392 45392 NUM bracis-28409 273 37 - - SYM bracis-28409 273 38 2 2 NUM bracis-28409 273 39 ebook ebook NOUN bracis-28409 273 40 packages package NOUN bracis-28409 273 41 : : PUNCT bracis-28409 273 42 computer computer NOUN bracis-28409 273 43 sciencecomputer sciencecomputer NOUN bracis-28409 273 44 science science NOUN bracis-28409 273 45 ( ( PUNCT bracis-28409 273 46 r0 r0 NOUN bracis-28409 273 47 ) ) PUNCT bracis-28409 273 48 share share VERB bracis-28409 273 49 this this DET bracis-28409 273 50 paper paper NOUN bracis-28409 273 51 anyone anyone PRON bracis-28409 273 52 you you PRON bracis-28409 273 53 share share VERB bracis-28409 273 54 the the DET bracis-28409 273 55 following follow VERB bracis-28409 273 56 link link NOUN bracis-28409 273 57 with with ADP bracis-28409 273 58 will will AUX bracis-28409 273 59 be be AUX bracis-28409 273 60 able able ADJ bracis-28409 273 61 to to PART bracis-28409 273 62 read read VERB bracis-28409 273 63 this this DET bracis-28409 273 64 content content NOUN bracis-28409 273 65 : : PUNCT bracis-28409 273 66 get get VERB bracis-28409 273 67 shareable shareable ADJ bracis-28409 273 68 linksorry linksorry NOUN bracis-28409 273 69 , , PUNCT bracis-28409 273 70 a a DET bracis-28409 273 71 shareable shareable ADJ bracis-28409 273 72 link link NOUN bracis-28409 273 73 is be AUX bracis-28409 273 74 not not PART bracis-28409 273 75 currently currently ADV bracis-28409 273 76 available available ADJ bracis-28409 273 77 for for ADP bracis-28409 273 78 this this DET bracis-28409 273 79 article article NOUN bracis-28409 273 80 . . PUNCT bracis-28409 274 1 copy copy VERB bracis-28409 274 2 shareable shareable ADJ bracis-28409 274 3 link link NOUN bracis-28409 274 4 to to PART bracis-28409 274 5 clipboard clipboard NOUN bracis-28409 274 6 provided provide VERB bracis-28409 274 7 by by ADP bracis-28409 274 8 the the DET bracis-28409 274 9 springer springer NOUN bracis-28409 274 10 nature nature PROPN bracis-28409 274 11 sharedit sharedit PROPN bracis-28409 274 12 content content NOUN bracis-28409 274 13 - - PUNCT bracis-28409 274 14 sharing share VERB bracis-28409 274 15 initiative initiative NOUN bracis-28409 274 16 keywords keyword NOUN bracis-28409 274 17 classifier classifier AUX bracis-28409 274 18 ensemble ensemble ADJ bracis-28409 274 19 self self NOUN bracis-28409 274 20 - - PUNCT bracis-28409 274 21 training training NOUN bracis-28409 274 22 algorithm algorithm NOUN bracis-28409 274 23 flexcon flexcon NOUN bracis-28409 274 24 - - PUNCT bracis-28409 274 25 c c NOUN bracis-28409 274 26 publish publish NOUN bracis-28409 274 27 with with ADP bracis-28409 274 28 us us PROPN bracis-28409 274 29 policies policy NOUN bracis-28409 274 30 and and CCONJ bracis-28409 274 31 ethics ethic NOUN bracis-28409 274 32 search search NOUN bracis-28409 274 33 search search NOUN bracis-28409 274 34 by by ADP bracis-28409 274 35 keyword keyword NOUN bracis-28409 274 36 or or CCONJ bracis-28409 274 37 author author NOUN bracis-28409 274 38 search search NOUN bracis-28409 274 39 navigation navigation NOUN bracis-28409 274 40 find find VERB bracis-28409 274 41 a a DET bracis-28409 274 42 journal journal NOUN bracis-28409 274 43 publish publish VERB bracis-28409 274 44 with with ADP bracis-28409 274 45 us we PRON bracis-28409 274 46 track track VERB bracis-28409 274 47 your your PRON bracis-28409 274 48 research research NOUN bracis-28409 274 49 discover discover VERB bracis-28409 274 50 content content NOUN bracis-28409 274 51 journals journal NOUN bracis-28409 274 52 a a DET bracis-28409 274 53 - - PUNCT bracis-28409 274 54 z z NOUN bracis-28409 274 55 books book NOUN bracis-28409 274 56 a a DET bracis-28409 274 57 - - PUNCT bracis-28409 274 58 z z NOUN bracis-28409 274 59 publish publish NOUN bracis-28409 274 60 with with ADP bracis-28409 274 61 us us PROPN bracis-28409 274 62 journal journal PROPN bracis-28409 274 63 finder finder PROPN bracis-28409 274 64 publish publish VERB bracis-28409 274 65 your your PRON bracis-28409 274 66 research research NOUN bracis-28409 274 67 language language NOUN bracis-28409 274 68 editing edit VERB bracis-28409 274 69 open open ADJ bracis-28409 274 70 access access NOUN bracis-28409 274 71 publishing publishing NOUN bracis-28409 274 72 products product NOUN bracis-28409 274 73 and and CCONJ bracis-28409 274 74 services service NOUN bracis-28409 274 75 our our PRON bracis-28409 274 76 products product NOUN bracis-28409 274 77 librarians librarian VERB bracis-28409 274 78 societies society NOUN bracis-28409 274 79 partners partner NOUN bracis-28409 274 80 and and CCONJ bracis-28409 274 81 advertisers advertiser NOUN bracis-28409 274 82 our our PRON bracis-28409 274 83 brands brand NOUN bracis-28409 274 84 springer springer NOUN bracis-28409 274 85 nature nature PROPN bracis-28409 274 86 portfolio portfolio PROPN bracis-28409 274 87 bmc bmc PROPN bracis-28409 274 88 palgrave palgrave PROPN bracis-28409 274 89 macmillan macmillan PROPN bracis-28409 274 90 apress apress PROPN bracis-28409 274 91 discover discover VERB bracis-28409 274 92 your your PRON bracis-28409 274 93 privacy privacy NOUN bracis-28409 274 94 choices choice NOUN bracis-28409 274 95 / / SYM bracis-28409 274 96 manage manage NOUN bracis-28409 274 97 cookies cookie NOUN bracis-28409 274 98 your your PRON bracis-28409 274 99 us us PROPN bracis-28409 275 1 state state NOUN bracis-28409 275 2 privacy privacy NOUN bracis-28409 275 3 rights right NOUN bracis-28409 275 4 accessibility accessibility NOUN bracis-28409 275 5 statement statement NOUN bracis-28409 275 6 terms term NOUN bracis-28409 275 7 and and CCONJ bracis-28409 275 8 conditions condition NOUN bracis-28409 275 9 privacy privacy NOUN bracis-28409 275 10 policy policy NOUN bracis-28409 275 11 help help NOUN bracis-28409 275 12 and and CCONJ bracis-28409 275 13 support support VERB bracis-28409 275 14 legal legal ADJ bracis-28409 275 15 notice notice NOUN bracis-28409 275 16 cancel cancel VERB bracis-28409 275 17 contracts contract NOUN bracis-28409 275 18 here here ADV bracis-28409 275 19 129.74.145.123 129.74.145.123 NUM bracis-28409 275 20 hesburgh hesburgh PROPN bracis-28409 275 21 library library PROPN bracis-28409 275 22 er er INTJ bracis-28409 275 23 unit unit NOUN bracis-28409 275 24 ( ( PUNCT bracis-28409 275 25 3005732405 3005732405 NUM bracis-28409 275 26 ) ) PUNCT bracis-28409 275 27 northeast northeast ADJ bracis-28409 275 28 research research NOUN bracis-28409 275 29 libraries library NOUN bracis-28409 275 30 ( ( PUNCT bracis-28409 275 31 nerl nerl PROPN bracis-28409 275 32 ) ) PUNCT bracis-28409 275 33 ( ( PUNCT bracis-28409 275 34 8200828607 8200828607 NUM bracis-28409 275 35 ) ) PUNCT bracis-28409 275 36 nerl nerl VERB bracis-28409 275 37 ta ta X bracis-28409 275 38 account account NOUN bracis-28409 275 39 ( ( PUNCT bracis-28409 275 40 3006206169 3006206169 NUM bracis-28409 275 41 ) ) PUNCT bracis-28409 275 42 university university NOUN bracis-28409 275 43 of of ADP bracis-28409 275 44 notre notre PROPN bracis-28409 275 45 dame dame PROPN bracis-28409 275 46 hesburgh hesburgh PROPN bracis-28409 275 47 library library NOUN bracis-28409 275 48 ( ( PUNCT bracis-28409 275 49 3000184373 3000184373 NUM bracis-28409 275 50 ) ) PUNCT bracis-28409 276 1 © © ADP bracis-28409 276 2 2025 2025 NUM bracis-28409 276 3 springer springer NOUN bracis-28409 276 4 nature nature NOUN