id sid tid token lemma pos bracis-28396 1 1 feature feature NOUN bracis-28396 1 2 selection selection NOUN bracis-28396 1 3 and and CCONJ bracis-28396 1 4     SPACE bracis-28396 1 5 hyperparameter hyperparameter NOUN bracis-28396 1 6 fine fine ADV bracis-28396 1 7 - - PUNCT bracis-28396 1 8 tuning tuning NOUN bracis-28396 1 9 in in ADP bracis-28396 1 10     SPACE bracis-28396 1 11 artificial artificial ADJ bracis-28396 1 12 neural neural ADJ bracis-28396 1 13 networks network NOUN bracis-28396 1 14 for for ADP bracis-28396 1 15     SPACE bracis-28396 1 16 wood wood NOUN bracis-28396 1 17 quality quality NOUN bracis-28396 1 18 classification classification NOUN bracis-28396 2 1 | | NOUN bracis-28396 2 2 springer springer NOUN bracis-28396 2 3 nature nature PROPN bracis-28396 2 4 link link PROPN bracis-28396 2 5 ( ( PUNCT bracis-28396 2 6 formerly formerly ADV bracis-28396 2 7 springerlink springerlink NOUN bracis-28396 2 8 ) ) PUNCT bracis-28396 2 9 skip skip VERB bracis-28396 2 10 to to ADP bracis-28396 2 11 main main ADJ bracis-28396 2 12 content content NOUN bracis-28396 2 13 advertisement advertisement NOUN bracis-28396 2 14 log log NOUN bracis-28396 2 15 in in ADP bracis-28396 2 16 menu menu NOUN bracis-28396 2 17 find find VERB bracis-28396 2 18 a a DET bracis-28396 2 19 journal journal NOUN bracis-28396 2 20 publish publish VERB bracis-28396 2 21 with with ADP bracis-28396 2 22 us we PRON bracis-28396 2 23 track track VERB bracis-28396 2 24 your your PRON bracis-28396 2 25 research research NOUN bracis-28396 2 26 search search NOUN bracis-28396 2 27 cart cart NOUN bracis-28396 2 28 home home NOUN bracis-28396 2 29 intelligent intelligent ADJ bracis-28396 2 30 systems system NOUN bracis-28396 2 31 conference conference NOUN bracis-28396 2 32 paper paper NOUN bracis-28396 2 33 feature feature NOUN bracis-28396 2 34 selection selection NOUN bracis-28396 2 35 and and CCONJ bracis-28396 2 36     SPACE bracis-28396 2 37 hyperparameter hyperparameter NOUN bracis-28396 2 38 fine fine ADV bracis-28396 2 39 - - PUNCT bracis-28396 2 40 tuning tuning NOUN bracis-28396 2 41 in in ADP bracis-28396 2 42     SPACE bracis-28396 2 43 artificial artificial ADJ bracis-28396 2 44 neural neural ADJ bracis-28396 2 45 networks network NOUN bracis-28396 2 46 for for ADP bracis-28396 2 47     SPACE bracis-28396 2 48 wood wood NOUN bracis-28396 2 49 quality quality NOUN bracis-28396 2 50 classification classification NOUN bracis-28396 2 51 conference conference NOUN bracis-28396 2 52 paper paper NOUN bracis-28396 2 53 first first ADV bracis-28396 2 54 online online ADV bracis-28396 2 55 : : PUNCT bracis-28396 2 56 12 12 NUM bracis-28396 2 57 october october NOUN bracis-28396 2 58 2023 2023 NUM bracis-28396 2 59 pp pp PROPN bracis-28396 2 60 323–337 323–337 NUM bracis-28396 2 61 cite cite VERB bracis-28396 2 62 this this DET bracis-28396 2 63 conference conference NOUN bracis-28396 2 64 paper paper NOUN bracis-28396 2 65 access access NOUN bracis-28396 2 66 provided provide VERB bracis-28396 2 67 by by ADP bracis-28396 2 68 university university PROPN bracis-28396 2 69 of of ADP bracis-28396 2 70 notre notre PROPN bracis-28396 2 71 dame dame PROPN bracis-28396 2 72 hesburgh hesburgh PROPN bracis-28396 2 73 library library PROPN bracis-28396 2 74 download download PROPN bracis-28396 2 75 book book NOUN bracis-28396 2 76 pdf pdf PROPN bracis-28396 2 77 download download NOUN bracis-28396 2 78 book book NOUN bracis-28396 2 79 epub epub PROPN bracis-28396 2 80 intelligent intelligent ADJ bracis-28396 2 81 systems system NOUN bracis-28396 2 82 ( ( PUNCT bracis-28396 2 83 bracis bracis NOUN bracis-28396 2 84 2023 2023 NUM bracis-28396 2 85 ) ) PUNCT bracis-28396 2 86 feature feature NOUN bracis-28396 2 87 selection selection NOUN bracis-28396 2 88 and and CCONJ bracis-28396 2 89     SPACE bracis-28396 2 90 hyperparameter hyperparameter NOUN bracis-28396 2 91 fine fine ADV bracis-28396 2 92 - - PUNCT bracis-28396 2 93 tuning tuning NOUN bracis-28396 2 94 in in ADP bracis-28396 2 95     SPACE bracis-28396 2 96 artificial artificial ADJ bracis-28396 2 97 neural neural ADJ bracis-28396 2 98 networks network NOUN bracis-28396 2 99 for for ADP bracis-28396 2 100     SPACE bracis-28396 2 101 wood wood NOUN bracis-28396 2 102 quality quality NOUN bracis-28396 2 103 classification classification NOUN bracis-28396 2 104 download download NOUN bracis-28396 2 105 book book NOUN bracis-28396 2 106 pdf pdf PROPN bracis-28396 2 107 download download NOUN bracis-28396 2 108 book book PROPN bracis-28396 2 109 epub epub PROPN bracis-28396 2 110 mateus mateus PROPN bracis-28396 2 111 roder roder PROPN bracis-28396 2 112     SPACE bracis-28396 2 113 orcid orcid NOUN bracis-28396 2 114 : : PUNCT bracis-28396 3 1 orcid.org/0000-0002-3112-52909 orcid.org/0000-0002-3112-52909 ADJ bracis-28396 3 2 , , PUNCT bracis-28396 3 3 leandro leandro PROPN bracis-28396 3 4 aparecido aparecido PROPN bracis-28396 3 5 passos passos PROPN bracis-28396 3 6     SPACE bracis-28396 3 7 orcid orcid NOUN bracis-28396 3 8 : : PUNCT bracis-28396 4 1 orcid.org/0000-0003-3529-31099 orcid.org/0000-0003-3529-31099 VERB bracis-28396 4 2 , , PUNCT bracis-28396 4 3 joão joão PROPN bracis-28396 4 4 paulo paulo PROPN bracis-28396 4 5 papa papa PROPN bracis-28396 4 6     SPACE bracis-28396 4 7 orcid orcid NOUN bracis-28396 4 8 : : PUNCT bracis-28396 4 9 orcid.org/0000-0002-6494-75149 orcid.org/0000-0002-6494-75149 PROPN bracis-28396 4 10 & & CCONJ bracis-28396 4 11 … … PUNCT bracis-28396 4 12 andré andré PROPN bracis-28396 4 13 luis luis PROPN bracis-28396 4 14 debiaso debiaso PROPN bracis-28396 4 15 rossi rossi PROPN bracis-28396 4 16     SPACE bracis-28396 4 17 orcid orcid NOUN bracis-28396 4 18 : : PUNCT bracis-28396 4 19 orcid.org/0000-0001-6388-747910 orcid.org/0000-0001-6388-747910 NOUN bracis-28396 4 20     SPACE bracis-28396 4 21 show show VERB bracis-28396 4 22 authors author NOUN bracis-28396 4 23 part part NOUN bracis-28396 4 24 of of ADP bracis-28396 4 25 the the DET bracis-28396 4 26 book book NOUN bracis-28396 4 27 series series NOUN bracis-28396 4 28 : : PUNCT bracis-28396 4 29 lecture lecture NOUN bracis-28396 4 30 notes note NOUN bracis-28396 4 31 in in ADP bracis-28396 4 32 computer computer NOUN bracis-28396 4 33 science science NOUN bracis-28396 4 34 ( ( PUNCT bracis-28396 4 35 ( ( PUNCT bracis-28396 4 36 lnai lnai ADJ bracis-28396 4 37 , , PUNCT bracis-28396 4 38 volume volume NOUN bracis-28396 4 39 14196 14196 NUM bracis-28396 4 40 ) ) PUNCT bracis-28396 4 41 ) ) PUNCT bracis-28396 4 42 included include VERB bracis-28396 4 43 in in ADP bracis-28396 4 44 the the DET bracis-28396 4 45 following follow VERB bracis-28396 4 46 conference conference NOUN bracis-28396 4 47 series series NOUN bracis-28396 4 48 : : PUNCT bracis-28396 4 49 brazilian brazilian ADJ bracis-28396 4 50 conference conference NOUN bracis-28396 4 51 on on ADP bracis-28396 4 52 intelligent intelligent ADJ bracis-28396 4 53 systems system NOUN bracis-28396 4 54 525 525 NUM bracis-28396 4 55 accesses access NOUN bracis-28396 4 56 1 1 NUM bracis-28396 4 57 citation citation NOUN bracis-28396 4 58 abstract abstract ADJ bracis-28396 4 59 quality quality NOUN bracis-28396 4 60 classification classification NOUN bracis-28396 4 61 of of ADP bracis-28396 4 62 wood wood NOUN bracis-28396 4 63 boards board NOUN bracis-28396 4 64 is be AUX bracis-28396 4 65 an an DET bracis-28396 4 66 essential essential ADJ bracis-28396 4 67 task task NOUN bracis-28396 4 68 in in ADP bracis-28396 4 69 the the DET bracis-28396 4 70 sawmill sawmill NOUN bracis-28396 4 71 industry industry NOUN bracis-28396 4 72 , , PUNCT bracis-28396 4 73 which which PRON bracis-28396 4 74 is be AUX bracis-28396 4 75 still still ADV bracis-28396 4 76 usually usually ADV bracis-28396 4 77 performed perform VERB bracis-28396 4 78 by by ADP bracis-28396 4 79 human human ADJ bracis-28396 4 80 operators operator NOUN bracis-28396 4 81 in in ADP bracis-28396 4 82 small small ADJ bracis-28396 4 83 to to ADP bracis-28396 5 1 median median ADJ bracis-28396 5 2 companies company NOUN bracis-28396 5 3 in in ADP bracis-28396 5 4 developing develop VERB bracis-28396 5 5 countries country NOUN bracis-28396 5 6 . . PUNCT bracis-28396 6 1 machine machine NOUN bracis-28396 6 2 learning learn VERB bracis-28396 6 3 algorithms algorithm NOUN bracis-28396 6 4 have have AUX bracis-28396 6 5 been be AUX bracis-28396 6 6 successfully successfully ADV bracis-28396 6 7 employed employ VERB bracis-28396 6 8 to to PART bracis-28396 6 9 investigate investigate VERB bracis-28396 6 10 the the DET bracis-28396 6 11 problem problem NOUN bracis-28396 6 12 , , PUNCT bracis-28396 6 13 offering offer VERB bracis-28396 6 14 a a DET bracis-28396 6 15 more more ADV bracis-28396 6 16 affordable affordable ADJ bracis-28396 6 17 alternative alternative NOUN bracis-28396 6 18 compared compare VERB bracis-28396 6 19 to to ADP bracis-28396 6 20 other other ADJ bracis-28396 6 21 solutions solution NOUN bracis-28396 6 22 . . PUNCT bracis-28396 7 1 however however ADV bracis-28396 7 2 , , PUNCT bracis-28396 7 3 such such ADJ bracis-28396 7 4 approaches approach NOUN bracis-28396 7 5 usually usually ADV bracis-28396 7 6 present present VERB bracis-28396 7 7 some some DET bracis-28396 7 8 drawbacks drawback NOUN bracis-28396 7 9 regarding regard VERB bracis-28396 7 10 the the DET bracis-28396 7 11 proper proper ADJ bracis-28396 7 12 selection selection NOUN bracis-28396 7 13 of of ADP bracis-28396 7 14 their their PRON bracis-28396 7 15 hyperparameters hyperparameter NOUN bracis-28396 7 16 . . PUNCT bracis-28396 8 1 moreover moreover ADV bracis-28396 8 2 , , PUNCT bracis-28396 8 3 the the DET bracis-28396 8 4 models model NOUN bracis-28396 8 5 are be AUX bracis-28396 8 6 susceptible susceptible ADJ bracis-28396 8 7 to to ADP bracis-28396 8 8 the the DET bracis-28396 8 9 features feature NOUN bracis-28396 8 10 extracted extract VERB bracis-28396 8 11 from from ADP bracis-28396 8 12 wood wood NOUN bracis-28396 8 13 board board NOUN bracis-28396 8 14 images image NOUN bracis-28396 8 15 , , PUNCT bracis-28396 8 16 which which PRON bracis-28396 8 17 influence influence VERB bracis-28396 8 18 the the DET bracis-28396 8 19 induction induction NOUN bracis-28396 8 20 of of ADP bracis-28396 8 21 the the DET bracis-28396 8 22 model model NOUN bracis-28396 8 23 and and CCONJ bracis-28396 8 24 , , PUNCT bracis-28396 8 25 consequently consequently ADV bracis-28396 8 26 , , PUNCT bracis-28396 8 27 its its PRON bracis-28396 8 28 generalization generalization NOUN bracis-28396 8 29 power power NOUN bracis-28396 8 30 . . PUNCT bracis-28396 9 1 therefore therefore ADV bracis-28396 9 2 , , PUNCT bracis-28396 9 3 in in ADP bracis-28396 9 4 this this DET bracis-28396 9 5 paper paper NOUN bracis-28396 9 6 , , PUNCT bracis-28396 9 7 we we PRON bracis-28396 9 8 investigate investigate VERB bracis-28396 9 9 the the DET bracis-28396 9 10 problem problem NOUN bracis-28396 9 11 of of ADP bracis-28396 9 12 simultaneously simultaneously ADV bracis-28396 9 13 tuning tune VERB bracis-28396 9 14 the the DET bracis-28396 9 15 hyperparameters hyperparameter NOUN bracis-28396 9 16 of of ADP bracis-28396 9 17 an an DET bracis-28396 9 18 artificial artificial ADJ bracis-28396 9 19 neural neural ADJ bracis-28396 9 20 network network NOUN bracis-28396 9 21 ( ( PUNCT bracis-28396 9 22 ann ann PROPN bracis-28396 9 23 ) ) PUNCT bracis-28396 9 24 as as ADV bracis-28396 9 25 well well ADV bracis-28396 9 26 as as ADP bracis-28396 9 27 selecting select VERB bracis-28396 9 28 a a DET bracis-28396 9 29 subset subset NOUN bracis-28396 9 30 of of ADP bracis-28396 9 31 characteristics characteristic NOUN bracis-28396 9 32 that that PRON bracis-28396 9 33 better well ADV bracis-28396 9 34 describes describe VERB bracis-28396 9 35 the the DET bracis-28396 9 36 wood wood NOUN bracis-28396 9 37 board board NOUN bracis-28396 9 38 quality quality NOUN bracis-28396 9 39 . . PUNCT bracis-28396 10 1 experiments experiment NOUN bracis-28396 10 2 were be AUX bracis-28396 10 3 conducted conduct VERB bracis-28396 10 4 over over ADP bracis-28396 10 5 a a DET bracis-28396 10 6 private private ADJ bracis-28396 10 7 dataset dataset NOUN bracis-28396 10 8 composed compose VERB bracis-28396 10 9 of of ADP bracis-28396 10 10 images image NOUN bracis-28396 10 11 obtained obtain VERB bracis-28396 10 12 from from ADP bracis-28396 10 13 a a DET bracis-28396 10 14 sawmill sawmill NOUN bracis-28396 10 15 industry industry NOUN bracis-28396 10 16 and and CCONJ bracis-28396 10 17 described describe VERB bracis-28396 10 18 using use VERB bracis-28396 10 19 different different ADJ bracis-28396 10 20 feature feature NOUN bracis-28396 10 21 descriptors descriptor NOUN bracis-28396 10 22 . . PUNCT bracis-28396 11 1 the the DET bracis-28396 11 2 predictive predictive ADJ bracis-28396 11 3 performance performance NOUN bracis-28396 11 4 of of ADP bracis-28396 11 5 the the DET bracis-28396 11 6 model model NOUN bracis-28396 11 7 was be AUX bracis-28396 11 8 compared compare VERB bracis-28396 11 9 against against ADP bracis-28396 11 10 five five NUM bracis-28396 11 11 baseline baseline ADJ bracis-28396 11 12 methods method NOUN bracis-28396 11 13 as as ADV bracis-28396 11 14 well well ADV bracis-28396 11 15 as as ADP bracis-28396 11 16 a a DET bracis-28396 11 17 random random ADJ bracis-28396 11 18 search search NOUN bracis-28396 11 19 , , PUNCT bracis-28396 11 20 performing perform VERB bracis-28396 11 21 either either CCONJ bracis-28396 11 22 ann ann PROPN bracis-28396 11 23 hyperparameter hyperparameter NOUN bracis-28396 11 24 tuning tuning NOUN bracis-28396 11 25 and and CCONJ bracis-28396 11 26 feature feature NOUN bracis-28396 11 27 selection selection NOUN bracis-28396 11 28 . . PUNCT bracis-28396 12 1 experimental experimental ADJ bracis-28396 12 2 results result NOUN bracis-28396 12 3 suggest suggest VERB bracis-28396 12 4 that that SCONJ bracis-28396 12 5 hyperparameters hyperparameter NOUN bracis-28396 12 6 should should AUX bracis-28396 12 7 be be AUX bracis-28396 12 8 adjusted adjust VERB bracis-28396 12 9 according accord VERB bracis-28396 12 10 to to ADP bracis-28396 12 11 the the DET bracis-28396 12 12 feature feature NOUN bracis-28396 12 13 set set NOUN bracis-28396 12 14 , , PUNCT bracis-28396 12 15 or or CCONJ bracis-28396 12 16 the the DET bracis-28396 12 17 features feature NOUN bracis-28396 12 18 should should AUX bracis-28396 12 19 be be AUX bracis-28396 12 20 selected select VERB bracis-28396 12 21 considering consider VERB bracis-28396 12 22 the the DET bracis-28396 12 23 hyperparameter hyperparameter NOUN bracis-28396 12 24 values value NOUN bracis-28396 12 25 . . PUNCT bracis-28396 13 1 in in ADP bracis-28396 13 2 summary summary NOUN bracis-28396 13 3 , , PUNCT bracis-28396 13 4 the the DET bracis-28396 13 5 best good ADJ bracis-28396 13 6 predictive predictive ADJ bracis-28396 13 7 performance performance NOUN bracis-28396 13 8 , , PUNCT bracis-28396 13 9 i.e. i.e. X bracis-28396 13 10 , , PUNCT bracis-28396 13 11 a a DET bracis-28396 13 12 balanced balanced ADJ bracis-28396 13 13 accuracy accuracy NOUN bracis-28396 13 14 of of ADP bracis-28396 13 15 0.80 0.80 NUM bracis-28396 13 16 , , PUNCT bracis-28396 13 17 was be AUX bracis-28396 13 18 achieved achieve VERB bracis-28396 13 19 in in ADP bracis-28396 13 20 two two NUM bracis-28396 13 21 distinct distinct ADJ bracis-28396 13 22 scenarios scenario NOUN bracis-28396 13 23 : : PUNCT bracis-28396 13 24 ( ( PUNCT bracis-28396 13 25 i i NOUN bracis-28396 13 26 ) ) PUNCT bracis-28396 13 27 performing perform VERB bracis-28396 13 28 only only ADJ bracis-28396 13 29 feature feature NOUN bracis-28396 13 30 selection selection NOUN bracis-28396 13 31 , , PUNCT bracis-28396 13 32 and and CCONJ bracis-28396 13 33 ( ( PUNCT bracis-28396 13 34 ii ii NOUN bracis-28396 13 35 ) ) PUNCT bracis-28396 13 36 performing perform VERB bracis-28396 13 37 both both DET bracis-28396 13 38 tasks task NOUN bracis-28396 13 39 concomitantly concomitantly ADV bracis-28396 13 40 . . PUNCT bracis-28396 14 1 thus thus ADV bracis-28396 14 2 , , PUNCT bracis-28396 14 3 we we PRON bracis-28396 14 4 suggest suggest VERB bracis-28396 14 5 that that SCONJ bracis-28396 14 6 at at ADV bracis-28396 14 7 least least ADJ bracis-28396 14 8 one one NUM bracis-28396 14 9 of of ADP bracis-28396 14 10 the the DET bracis-28396 14 11 two two NUM bracis-28396 14 12 approaches approach NOUN bracis-28396 14 13 should should AUX bracis-28396 14 14 be be AUX bracis-28396 14 15 considered consider VERB bracis-28396 14 16 in in ADP bracis-28396 14 17 the the DET bracis-28396 14 18 context context NOUN bracis-28396 14 19 of of ADP bracis-28396 14 20 industrial industrial ADJ bracis-28396 14 21 applications application NOUN bracis-28396 14 22 . . PUNCT bracis-28396 15 1 the the DET bracis-28396 15 2 authors author NOUN bracis-28396 15 3 are be AUX bracis-28396 15 4 grateful grateful ADJ bracis-28396 15 5 to to ADP bracis-28396 15 6 fapesp fapesp ADJ bracis-28396 15 7 grants grant NOUN bracis-28396 15 8 # # SYM bracis-28396 15 9 2016/06538 2016/06538 NUM bracis-28396 15 10 - - SYM bracis-28396 15 11 0 0 NUM bracis-28396 15 12 , , PUNCT bracis-28396 15 13 # # SYM bracis-28396 15 14 2018/02822 2018/02822 NUM bracis-28396 15 15 - - SYM bracis-28396 15 16 1 1 NUM bracis-28396 15 17 and and CCONJ bracis-28396 15 18 # # SYM bracis-28396 15 19 2019/07825 2019/07825 NUM bracis-28396 15 20 - - SYM bracis-28396 15 21 1 1 NUM bracis-28396 15 22 . . PUNCT bracis-28396 15 23 access access NOUN bracis-28396 15 24 provided provide VERB bracis-28396 15 25 by by ADP bracis-28396 15 26 university university PROPN bracis-28396 15 27 of of ADP bracis-28396 15 28 notre notre PROPN bracis-28396 15 29 dame dame PROPN bracis-28396 15 30 hesburgh hesburgh PROPN bracis-28396 15 31 library library PROPN bracis-28396 15 32 . . PUNCT bracis-28396 16 1 download download PROPN bracis-28396 16 2 conference conference NOUN bracis-28396 16 3 paper paper NOUN bracis-28396 16 4 pdf pdf NOUN bracis-28396 16 5 similar similar ADJ bracis-28396 16 6 content content NOUN bracis-28396 16 7 being be AUX bracis-28396 16 8 viewed view VERB bracis-28396 16 9 by by ADP bracis-28396 16 10 others other NOUN bracis-28396 16 11 artificial artificial ADJ bracis-28396 16 12 intelligence intelligence NOUN bracis-28396 16 13 - - PUNCT bracis-28396 16 14 driven drive VERB bracis-28396 16 15 timber timber NOUN bracis-28396 16 16 wood wood NOUN bracis-28396 16 17 defect defect NOUN bracis-28396 16 18 characterization characterization NOUN bracis-28396 16 19 from from ADP bracis-28396 16 20 terahertz terahertz NOUN bracis-28396 16 21 images image NOUN bracis-28396 16 22 article article NOUN bracis-28396 16 23 13 13 NUM bracis-28396 16 24 october october PROPN bracis-28396 16 25 2024 2024 NUM bracis-28396 16 26 employing employ VERB bracis-28396 16 27 artificial artificial ADJ bracis-28396 16 28 neural neural ADJ bracis-28396 16 29 networks network NOUN bracis-28396 16 30 for for ADP bracis-28396 16 31 minimizing minimize VERB bracis-28396 16 32 surface surface NOUN bracis-28396 16 33 roughness roughness NOUN bracis-28396 16 34 and and CCONJ bracis-28396 16 35 power power NOUN bracis-28396 16 36 consumption consumption NOUN bracis-28396 16 37 in in ADP bracis-28396 16 38 abrasive abrasive ADJ bracis-28396 16 39 machining machining NOUN bracis-28396 16 40 of of ADP bracis-28396 16 41 wood wood NOUN bracis-28396 16 42 article article NOUN bracis-28396 16 43 20 20 NUM bracis-28396 16 44 april april PROPN bracis-28396 16 45 2016 2016 NUM bracis-28396 16 46 production production NOUN bracis-28396 16 47 of of ADP bracis-28396 16 48 high high ADJ bracis-28396 16 49 - - PUNCT bracis-28396 16 50 quality quality NOUN bracis-28396 16 51 forest forest NOUN bracis-28396 16 52 wood wood NOUN bracis-28396 16 53 biomass biomass NOUN bracis-28396 16 54 using use VERB bracis-28396 16 55 artificial artificial ADJ bracis-28396 16 56 intelligence intelligence NOUN bracis-28396 16 57 to to PART bracis-28396 16 58 control control VERB bracis-28396 16 59 thermal thermal ADJ bracis-28396 16 60 modification modification NOUN bracis-28396 16 61 article article NOUN bracis-28396 16 62 26 26 NUM bracis-28396 16 63 april april PROPN bracis-28396 16 64 2022 2022 NUM bracis-28396 16 65 explore explore VERB bracis-28396 16 66 related relate VERB bracis-28396 16 67 subjects subject NOUN bracis-28396 16 68 discover discover VERB bracis-28396 16 69 the the DET bracis-28396 16 70 latest late ADJ bracis-28396 16 71 articles article NOUN bracis-28396 16 72 , , PUNCT bracis-28396 16 73 books book NOUN bracis-28396 16 74 and and CCONJ bracis-28396 16 75 news news NOUN bracis-28396 16 76 in in ADP bracis-28396 16 77 related related ADJ bracis-28396 16 78 subjects subject NOUN bracis-28396 16 79 , , PUNCT bracis-28396 16 80 suggested suggest VERB bracis-28396 16 81 using use VERB bracis-28396 16 82 machine machine NOUN bracis-28396 16 83 learning learning NOUN bracis-28396 16 84 . . PUNCT bracis-28396 17 1 categorization categorization NOUN bracis-28396 17 2 learning learn VERB bracis-28396 17 3 algorithms algorithm NOUN bracis-28396 17 4 machine machine NOUN bracis-28396 17 5 learning learn VERB bracis-28396 17 6 optimization optimization NOUN bracis-28396 17 7 statistical statistical ADJ bracis-28396 17 8 learning learning NOUN bracis-28396 17 9 wood wood NOUN bracis-28396 17 10 science science NOUN bracis-28396 17 11 and and CCONJ bracis-28396 17 12 technology technology NOUN bracis-28396 17 13 1 1 NUM bracis-28396 17 14 introduction introduction NOUN bracis-28396 17 15 industries industry NOUN bracis-28396 17 16 have have AUX bracis-28396 17 17 experienced experience VERB bracis-28396 17 18 many many ADJ bracis-28396 17 19 technological technological ADJ bracis-28396 17 20 advances advance NOUN bracis-28396 17 21 in in ADP bracis-28396 17 22 recent recent ADJ bracis-28396 17 23 years year NOUN bracis-28396 17 24 , , PUNCT bracis-28396 17 25 resulting result VERB bracis-28396 17 26 in in ADP bracis-28396 17 27 more more ADJ bracis-28396 17 28 complex complex ADJ bracis-28396 17 29 processes process NOUN bracis-28396 17 30 , , PUNCT bracis-28396 17 31 systems system NOUN bracis-28396 17 32 , , PUNCT bracis-28396 17 33 and and CCONJ bracis-28396 17 34 products product NOUN bracis-28396 17 35 . . PUNCT bracis-28396 18 1 as as ADP bracis-28396 18 2 a a DET bracis-28396 18 3 consequence consequence NOUN bracis-28396 18 4 , , PUNCT bracis-28396 18 5 the the DET bracis-28396 18 6 management management NOUN bracis-28396 18 7 of of ADP bracis-28396 18 8 integrated integrate VERB bracis-28396 18 9 manufacturing manufacturing NOUN bracis-28396 18 10 processes process NOUN bracis-28396 18 11 and and CCONJ bracis-28396 18 12 operation operation NOUN bracis-28396 18 13 analyses analysis NOUN bracis-28396 18 14 are be AUX bracis-28396 18 15 crucial crucial ADJ bracis-28396 18 16 to to ADP bracis-28396 18 17 delivering deliver VERB bracis-28396 18 18 high high ADJ bracis-28396 18 19 - - PUNCT bracis-28396 18 20 quality quality NOUN bracis-28396 18 21 products product NOUN bracis-28396 18 22 to to ADP bracis-28396 18 23 clients client NOUN bracis-28396 18 24 . . PUNCT bracis-28396 19 1 in in ADP bracis-28396 19 2 this this DET bracis-28396 19 3 scenario scenario NOUN bracis-28396 19 4 , , PUNCT bracis-28396 19 5 the the DET bracis-28396 19 6 quality quality NOUN bracis-28396 19 7 of of ADP bracis-28396 19 8 raw raw ADJ bracis-28396 19 9 materials material NOUN bracis-28396 19 10 is be AUX bracis-28396 19 11 also also ADV bracis-28396 19 12 paramount paramount ADJ bracis-28396 19 13 for for ADP bracis-28396 19 14 high high ADJ bracis-28396 19 15 quality quality NOUN bracis-28396 19 16 manufactured manufacture VERB bracis-28396 19 17 products product NOUN bracis-28396 19 18 . . PUNCT bracis-28396 20 1 however however ADV bracis-28396 20 2 , , PUNCT bracis-28396 20 3 imperfection imperfection NOUN bracis-28396 20 4 detection detection NOUN bracis-28396 20 5 , , PUNCT bracis-28396 20 6 as as ADV bracis-28396 20 7 well well ADV bracis-28396 20 8 as as ADP bracis-28396 20 9 quality quality NOUN bracis-28396 20 10 classification classification NOUN bracis-28396 20 11 of of ADP bracis-28396 20 12 raw raw ADJ bracis-28396 20 13 materials material NOUN bracis-28396 20 14 , , PUNCT bracis-28396 20 15 such such ADJ bracis-28396 20 16 as as ADP bracis-28396 20 17 woods wood NOUN bracis-28396 20 18 in in ADP bracis-28396 20 19 sawmill sawmill NOUN bracis-28396 20 20 companies company NOUN bracis-28396 20 21 , , PUNCT bracis-28396 20 22 usually usually ADV bracis-28396 20 23 is be AUX bracis-28396 20 24 still still ADV bracis-28396 20 25 performed perform VERB bracis-28396 20 26 by by ADP bracis-28396 20 27 trained train VERB bracis-28396 20 28 human human ADJ bracis-28396 20 29 operators operator NOUN bracis-28396 20 30     SPACE bracis-28396 21 1 [ [ X bracis-28396 21 2 1 1 NUM bracis-28396 21 3 ] ] PUNCT bracis-28396 21 4 . . PUNCT bracis-28396 22 1 notwithstanding notwithstanding ADV bracis-28396 22 2 , , PUNCT bracis-28396 22 3 the the DET bracis-28396 22 4 process process NOUN bracis-28396 22 5 is be AUX bracis-28396 22 6 inherently inherently ADV bracis-28396 22 7 subjective subjective ADJ bracis-28396 22 8 , , PUNCT bracis-28396 22 9 since since SCONJ bracis-28396 22 10 it it PRON bracis-28396 22 11 is be AUX bracis-28396 22 12 a a DET bracis-28396 22 13 visual visual ADJ bracis-28396 22 14 analysis analysis NOUN bracis-28396 22 15 , , PUNCT bracis-28396 22 16 and and CCONJ bracis-28396 22 17 these these DET bracis-28396 22 18 experts expert NOUN bracis-28396 22 19 may may AUX bracis-28396 22 20 suffer suffer VERB bracis-28396 22 21 from from ADP bracis-28396 22 22 fatigue fatigue NOUN bracis-28396 22 23 after after ADP bracis-28396 22 24 a a DET bracis-28396 22 25 long long ADJ bracis-28396 22 26 working working NOUN bracis-28396 22 27 period period NOUN bracis-28396 22 28 performing perform VERB bracis-28396 22 29 repetitive repetitive ADJ bracis-28396 22 30 activities activity NOUN bracis-28396 22 31 . . PUNCT bracis-28396 23 1 consequently consequently ADV bracis-28396 23 2 , , PUNCT bracis-28396 23 3 it it PRON bracis-28396 23 4 is be AUX bracis-28396 23 5 expected expect VERB bracis-28396 23 6 the the DET bracis-28396 23 7 increased increase VERB bracis-28396 23 8 number number NOUN bracis-28396 23 9 of of ADP bracis-28396 23 10 incorrect incorrect ADJ bracis-28396 23 11 classifications classification NOUN bracis-28396 23 12     SPACE bracis-28396 23 13 [ [ X bracis-28396 23 14 25 25 NUM bracis-28396 23 15 ] ] PUNCT bracis-28396 23 16 . . PUNCT bracis-28396 24 1 these these DET bracis-28396 24 2 disadvantages disadvantage NOUN bracis-28396 24 3 stimulated stimulate VERB bracis-28396 24 4 the the DET bracis-28396 24 5 scientific scientific ADJ bracis-28396 24 6 community community NOUN bracis-28396 24 7 towards towards ADP bracis-28396 24 8 the the DET bracis-28396 24 9 implementation implementation NOUN bracis-28396 24 10 of of ADP bracis-28396 24 11 visual visual ADJ bracis-28396 24 12 inspection inspection NOUN bracis-28396 24 13 systems system NOUN bracis-28396 24 14 , , PUNCT bracis-28396 24 15 aiming aim VERB bracis-28396 24 16 to to PART bracis-28396 24 17 perform perform VERB bracis-28396 24 18 defect defect NOUN bracis-28396 24 19 and and CCONJ bracis-28396 24 20 quality quality NOUN bracis-28396 24 21 classification classification NOUN bracis-28396 24 22 autonomously autonomously ADV bracis-28396 24 23 . . PUNCT bracis-28396 25 1 therefore therefore ADV bracis-28396 25 2 , , PUNCT bracis-28396 25 3 this this DET bracis-28396 25 4 work work NOUN bracis-28396 25 5 focuses focus VERB bracis-28396 25 6 exclusively exclusively ADV bracis-28396 25 7 on on ADP bracis-28396 25 8 machine machine NOUN bracis-28396 25 9 learning learning NOUN bracis-28396 25 10 ( ( PUNCT bracis-28396 25 11 ml ml NOUN bracis-28396 25 12 ) ) PUNCT bracis-28396 25 13 techniques technique NOUN bracis-28396 25 14 that that PRON bracis-28396 25 15 have have AUX bracis-28396 25 16 been be AUX bracis-28396 25 17 successfully successfully ADV bracis-28396 25 18 employed employ VERB bracis-28396 25 19 for for ADP bracis-28396 25 20 these these DET bracis-28396 25 21 tasks task NOUN bracis-28396 25 22 on on ADP bracis-28396 25 23 wood wood NOUN bracis-28396 25 24 boards board NOUN bracis-28396 25 25     SPACE bracis-28396 26 1 [ [ X bracis-28396 26 2 6 6 NUM bracis-28396 26 3 ] ] PUNCT bracis-28396 26 4 . . PUNCT bracis-28396 27 1 recently recently ADV bracis-28396 27 2 , , PUNCT bracis-28396 27 3 some some DET bracis-28396 27 4 studies study NOUN bracis-28396 27 5 investigated investigate VERB bracis-28396 27 6 the the DET bracis-28396 27 7 performance performance NOUN bracis-28396 27 8 of of ADP bracis-28396 27 9 different different ADJ bracis-28396 27 10 ml ml NOUN bracis-28396 27 11 techniques technique NOUN bracis-28396 27 12 to to PART bracis-28396 27 13 classify classify VERB bracis-28396 27 14 the the DET bracis-28396 27 15 quality quality NOUN bracis-28396 27 16 of of ADP bracis-28396 27 17 wood wood NOUN bracis-28396 27 18 surface surface NOUN bracis-28396 27 19     SPACE bracis-28396 28 1 [ [ X bracis-28396 28 2 17 17 NUM bracis-28396 28 3 , , PUNCT bracis-28396 28 4 21 21 NUM bracis-28396 28 5 ] ] PUNCT bracis-28396 28 6 . . PUNCT bracis-28396 29 1 these these DET bracis-28396 29 2 studies study NOUN bracis-28396 29 3 have have AUX bracis-28396 29 4 used use VERB bracis-28396 29 5 data datum NOUN bracis-28396 29 6 generated generate VERB bracis-28396 29 7 from from ADP bracis-28396 29 8 images image NOUN bracis-28396 29 9 captured capture VERB bracis-28396 29 10 in in ADP bracis-28396 29 11 a a DET bracis-28396 29 12 real real ADJ bracis-28396 29 13 sawmill sawmill NOUN bracis-28396 29 14 company company NOUN bracis-28396 29 15 and and CCONJ bracis-28396 29 16 classified classify VERB bracis-28396 29 17 by by ADP bracis-28396 29 18 a a DET bracis-28396 29 19 specialist specialist NOUN bracis-28396 29 20 in in ADP bracis-28396 29 21 three three NUM bracis-28396 29 22 levels level NOUN bracis-28396 29 23 of of ADP bracis-28396 29 24 quality quality NOUN bracis-28396 29 25 , , PUNCT bracis-28396 29 26 according accord VERB bracis-28396 29 27 to to ADP bracis-28396 29 28 the the DET bracis-28396 29 29 company company NOUN bracis-28396 29 30 rules rule NOUN bracis-28396 29 31 , , PUNCT bracis-28396 29 32 i.e. i.e. X bracis-28396 29 33 , , PUNCT bracis-28396 29 34 zero zero NUM bracis-28396 29 35 defect defect NOUN bracis-28396 29 36 is be AUX bracis-28396 29 37 found find VERB bracis-28396 29 38 in in ADP bracis-28396 29 39 the the DET bracis-28396 29 40 wood wood NOUN bracis-28396 29 41 piece piece NOUN bracis-28396 29 42 ( ( PUNCT bracis-28396 29 43 a a X bracis-28396 29 44 ) ) PUNCT bracis-28396 29 45 , , PUNCT bracis-28396 29 46 only only ADV bracis-28396 29 47 small small ADJ bracis-28396 29 48 defects defect NOUN bracis-28396 29 49 , , PUNCT bracis-28396 29 50 such such ADJ bracis-28396 29 51 as as ADP bracis-28396 29 52 knots knot NOUN bracis-28396 29 53 , , PUNCT bracis-28396 29 54 are be AUX bracis-28396 29 55 found find VERB bracis-28396 29 56 ( ( PUNCT bracis-28396 29 57 b b NOUN bracis-28396 29 58 ) ) PUNCT bracis-28396 29 59 , , PUNCT bracis-28396 29 60 and and CCONJ bracis-28396 29 61 defects defect NOUN bracis-28396 29 62 that that PRON bracis-28396 29 63 compromise compromise VERB bracis-28396 29 64 the the DET bracis-28396 29 65 quality quality NOUN bracis-28396 29 66 of of ADP bracis-28396 29 67 the the DET bracis-28396 29 68 product product NOUN bracis-28396 29 69 , , PUNCT bracis-28396 29 70 such such ADJ bracis-28396 29 71 as as ADP bracis-28396 29 72 groups group NOUN bracis-28396 29 73 of of ADP bracis-28396 29 74 knots knot NOUN bracis-28396 29 75 and and CCONJ bracis-28396 29 76 exposed expose VERB bracis-28396 29 77 pith pith NOUN bracis-28396 29 78 ( ( PUNCT bracis-28396 29 79 c c NOUN bracis-28396 29 80 ) ) PUNCT bracis-28396 29 81 . . PUNCT bracis-28396 30 1 figure figure VERB bracis-28396 30 2     SPACE bracis-28396 30 3 1 1 NUM bracis-28396 30 4 depicts depict VERB bracis-28396 30 5 some some DET bracis-28396 30 6 examples example NOUN bracis-28396 30 7 of of ADP bracis-28396 30 8 wood wood NOUN bracis-28396 30 9 images image NOUN bracis-28396 30 10 classified classify VERB bracis-28396 30 11 at at ADP bracis-28396 30 12 each each DET bracis-28396 30 13 level level NOUN bracis-28396 30 14 . . PUNCT bracis-28396 31 1 fig fig NOUN bracis-28396 31 2 . . PUNCT bracis-28396 32 1 1 1 NUM bracis-28396 32 2 . . X bracis-28396 32 3 three three NUM bracis-28396 32 4 different different ADJ bracis-28396 32 5 qualities quality NOUN bracis-28396 32 6 of of ADP bracis-28396 32 7 wood wood NOUN bracis-28396 32 8 boards board NOUN bracis-28396 32 9 ( ( PUNCT bracis-28396 32 10 a a DET bracis-28396 32 11 , , PUNCT bracis-28396 32 12 b b NOUN bracis-28396 32 13 , , PUNCT bracis-28396 32 14 and and CCONJ bracis-28396 32 15 c c X bracis-28396 32 16 ) ) PUNCT bracis-28396 32 17 according accord VERB bracis-28396 32 18 to to ADP bracis-28396 32 19 company company NOUN bracis-28396 32 20 ’s ’s PART bracis-28396 32 21 rule rule NOUN bracis-28396 32 22 . . PUNCT bracis-28396 33 1 full full ADJ bracis-28396 33 2 size size NOUN bracis-28396 33 3 image image NOUN bracis-28396 33 4 despite despite SCONJ bracis-28396 33 5 the the DET bracis-28396 33 6 success success NOUN bracis-28396 33 7 obtained obtain VERB bracis-28396 33 8 in in ADP bracis-28396 33 9 these these DET bracis-28396 33 10 studies study NOUN bracis-28396 33 11 , , PUNCT bracis-28396 33 12 each each DET bracis-28396 33 13 ml ml NOUN bracis-28396 33 14 algorithm algorithm PROPN bracis-28396 33 15 has have VERB bracis-28396 33 16 its its PRON bracis-28396 33 17 inherent inherent ADJ bracis-28396 33 18 tendency tendency NOUN bracis-28396 33 19 towards towards ADP bracis-28396 33 20 data datum NOUN bracis-28396 33 21 specificity specificity NOUN bracis-28396 33 22 , , PUNCT bracis-28396 33 23 which which PRON bracis-28396 33 24 influences influence VERB bracis-28396 33 25 the the DET bracis-28396 33 26 model model NOUN bracis-28396 33 27 ’s ’s PART bracis-28396 33 28 induction induction NOUN bracis-28396 33 29 and and CCONJ bracis-28396 33 30 , , PUNCT bracis-28396 33 31 thus thus ADV bracis-28396 33 32 , , PUNCT bracis-28396 33 33 its its PRON bracis-28396 33 34 predictive predictive ADJ bracis-28396 33 35 performance performance NOUN bracis-28396 33 36 . . PUNCT bracis-28396 34 1 therefore therefore ADV bracis-28396 34 2 , , PUNCT bracis-28396 34 3 one one PRON bracis-28396 34 4 can can AUX bracis-28396 34 5 adjust adjust VERB bracis-28396 34 6 such such ADJ bracis-28396 34 7 tendencies tendency NOUN bracis-28396 34 8 through through ADP bracis-28396 34 9 a a DET bracis-28396 34 10 proper proper ADJ bracis-28396 34 11 hyperparameter hyperparameter NOUN bracis-28396 34 12 ( ( PUNCT bracis-28396 34 13 hp hp NOUN bracis-28396 34 14 ) ) PUNCT bracis-28396 34 15 selection selection NOUN bracis-28396 34 16 . . PUNCT bracis-28396 35 1 the the DET bracis-28396 35 2 task task NOUN bracis-28396 35 3 of of ADP bracis-28396 35 4 finding find VERB bracis-28396 35 5 the the DET bracis-28396 35 6 best good ADJ bracis-28396 35 7 hp hp ADJ bracis-28396 35 8 values value NOUN bracis-28396 35 9 is be AUX bracis-28396 35 10 known know VERB bracis-28396 35 11 as as ADP bracis-28396 35 12 hyperparameter hyperparameter NOUN bracis-28396 35 13 tuning tuning NOUN bracis-28396 35 14 and and CCONJ bracis-28396 35 15 usually usually ADV bracis-28396 35 16 aims aim VERB bracis-28396 35 17 at at ADP bracis-28396 35 18 improving improve VERB bracis-28396 35 19 the the DET bracis-28396 35 20 model model NOUN bracis-28396 35 21 ’s ’s PART bracis-28396 35 22 predictive predictive ADJ bracis-28396 35 23 performance performance NOUN bracis-28396 35 24 while while SCONJ bracis-28396 35 25 keeping keep VERB bracis-28396 35 26 the the DET bracis-28396 35 27 model model NOUN bracis-28396 35 28 as as ADV bracis-28396 35 29 simple simple ADJ bracis-28396 35 30 as as ADP bracis-28396 35 31 possible possible ADJ bracis-28396 35 32 . . PUNCT bracis-28396 36 1 although although SCONJ bracis-28396 36 2 some some DET bracis-28396 36 3 hp hp ADJ bracis-28396 36 4 values value NOUN bracis-28396 36 5 may may AUX bracis-28396 36 6 fit fit VERB bracis-28396 36 7 sufficiently sufficiently ADV bracis-28396 36 8 well well ADV bracis-28396 36 9 different different ADJ bracis-28396 36 10 kinds kind NOUN bracis-28396 36 11 of of ADP bracis-28396 36 12 problems problem NOUN bracis-28396 36 13 , , PUNCT bracis-28396 36 14 it it PRON bracis-28396 36 15 is be AUX bracis-28396 36 16 a a DET bracis-28396 36 17 common common ADJ bracis-28396 36 18 practice practice NOUN bracis-28396 36 19 to to PART bracis-28396 36 20 search search VERB bracis-28396 36 21 for for ADP bracis-28396 36 22 the the DET bracis-28396 36 23 hyperparameter hyperparameter NOUN bracis-28396 36 24 that that PRON bracis-28396 36 25 provides provide VERB bracis-28396 36 26 the the DET bracis-28396 36 27 best good ADJ bracis-28396 36 28 solutions solution NOUN bracis-28396 36 29 concerning concern VERB bracis-28396 36 30 each each DET bracis-28396 36 31 problem problem NOUN bracis-28396 36 32 at at ADP bracis-28396 36 33 hand hand NOUN bracis-28396 36 34 [ [ X bracis-28396 36 35 17 17 NUM bracis-28396 36 36 , , PUNCT bracis-28396 36 37 21 21 NUM bracis-28396 36 38 ] ] PUNCT bracis-28396 36 39 . . PUNCT bracis-28396 37 1 besides besides SCONJ bracis-28396 37 2 , , PUNCT bracis-28396 37 3 another another DET bracis-28396 37 4 challenge challenge NOUN bracis-28396 37 5 in in ADP bracis-28396 37 6 the the DET bracis-28396 37 7 context context NOUN bracis-28396 37 8 of of ADP bracis-28396 37 9 this this DET bracis-28396 37 10 work work NOUN bracis-28396 37 11 is be AUX bracis-28396 37 12 extracting extract VERB bracis-28396 37 13 the the DET bracis-28396 37 14 images image NOUN bracis-28396 37 15 ’ ' PUNCT bracis-28396 37 16 more more ADJ bracis-28396 37 17 representative representative ADJ bracis-28396 37 18 features feature NOUN bracis-28396 37 19 , , PUNCT bracis-28396 37 20 i.e. i.e. X bracis-28396 37 21 , , PUNCT bracis-28396 37 22 the the DET bracis-28396 37 23 features feature NOUN bracis-28396 37 24 that that PRON bracis-28396 37 25 best good ADJ bracis-28396 37 26 describes describe VERB bracis-28396 37 27 the the DET bracis-28396 37 28 problem problem NOUN bracis-28396 37 29 , , PUNCT bracis-28396 37 30 since since SCONJ bracis-28396 37 31 the the DET bracis-28396 37 32 more more ADV bracis-28396 37 33 descriptive descriptive ADJ bracis-28396 37 34 they they PRON bracis-28396 37 35 are be AUX bracis-28396 37 36 , , PUNCT bracis-28396 37 37 the the PRON bracis-28396 37 38 higher high ADJ bracis-28396 37 39 the the DET bracis-28396 37 40 effectiveness effectiveness NOUN bracis-28396 37 41 of of ADP bracis-28396 37 42 the the DET bracis-28396 37 43 technique technique NOUN bracis-28396 37 44 . . PUNCT bracis-28396 38 1 regarding regard VERB bracis-28396 38 2 classification classification NOUN bracis-28396 38 3 tasks task NOUN bracis-28396 38 4 , , PUNCT bracis-28396 38 5 features feature NOUN bracis-28396 38 6 are be AUX bracis-28396 38 7 usually usually ADV bracis-28396 38 8 extracted extract VERB bracis-28396 38 9 through through ADP bracis-28396 38 10 image image NOUN bracis-28396 38 11 descriptors descriptor NOUN bracis-28396 38 12 , , PUNCT bracis-28396 38 13 such such ADJ bracis-28396 38 14 as as ADP bracis-28396 38 15 statistical statistical ADJ bracis-28396 38 16 measures measure NOUN bracis-28396 38 17 from from ADP bracis-28396 38 18 the the DET bracis-28396 38 19 gray gray ADJ bracis-28396 38 20 level level NOUN bracis-28396 38 21 co co NOUN bracis-28396 38 22 - - NOUN bracis-28396 38 23 occurrence occurrence ADJ bracis-28396 38 24 matrix matrix NOUN bracis-28396 38 25 ( ( PUNCT bracis-28396 38 26 glcm glcm PROPN bracis-28396 38 27 ) ) PUNCT bracis-28396 39 1 [ [ X bracis-28396 39 2 8 8 NUM bracis-28396 39 3 ] ] PUNCT bracis-28396 39 4 and and CCONJ bracis-28396 39 5 local local ADJ bracis-28396 39 6 binary binary ADJ bracis-28396 39 7 patterns pattern NOUN bracis-28396 39 8 ( ( PUNCT bracis-28396 39 9 lbp lbp PROPN bracis-28396 39 10 ) ) PUNCT bracis-28396 39 11     SPACE bracis-28396 40 1 [ [ X bracis-28396 40 2 13 13 NUM bracis-28396 40 3 ] ] PUNCT bracis-28396 40 4 . . PUNCT bracis-28396 41 1 however however ADV bracis-28396 41 2 , , PUNCT bracis-28396 41 3 many many ADJ bracis-28396 41 4 of of ADP bracis-28396 41 5 these these DET bracis-28396 41 6 features feature NOUN bracis-28396 41 7 may may AUX bracis-28396 41 8 be be AUX bracis-28396 41 9 correlated correlate VERB bracis-28396 41 10 to to ADP bracis-28396 41 11 each each DET bracis-28396 41 12 other other ADJ bracis-28396 41 13 or or CCONJ bracis-28396 41 14 do do AUX bracis-28396 41 15 not not PART bracis-28396 41 16 add add VERB bracis-28396 41 17 any any DET bracis-28396 41 18 relevant relevant ADJ bracis-28396 41 19 information information NOUN bracis-28396 41 20 for for ADP bracis-28396 41 21 the the DET bracis-28396 41 22 ml ml NOUN bracis-28396 41 23 technique technique NOUN bracis-28396 41 24 . . PUNCT bracis-28396 42 1 therefore therefore ADV bracis-28396 42 2 , , PUNCT bracis-28396 42 3 the the DET bracis-28396 42 4 process process NOUN bracis-28396 42 5 of of ADP bracis-28396 42 6 selecting select VERB bracis-28396 42 7 a a DET bracis-28396 42 8 subset subset NOUN bracis-28396 42 9 of of ADP bracis-28396 42 10 these these DET bracis-28396 42 11 features feature NOUN bracis-28396 42 12 , , PUNCT bracis-28396 42 13 referred refer VERB bracis-28396 42 14 to to ADP bracis-28396 42 15 as as ADP bracis-28396 42 16 feature feature NOUN bracis-28396 42 17 selection selection NOUN bracis-28396 42 18 ( ( PUNCT bracis-28396 42 19 fs fs PROPN bracis-28396 42 20 ) ) PUNCT bracis-28396 42 21 , , PUNCT bracis-28396 42 22 can can AUX bracis-28396 42 23 be be AUX bracis-28396 42 24 applied apply VERB bracis-28396 42 25 to to PART bracis-28396 42 26 select select VERB bracis-28396 42 27 the the DET bracis-28396 42 28 most most ADV bracis-28396 42 29 descriptive descriptive ADJ bracis-28396 42 30 ones one NOUN bracis-28396 42 31 . . PUNCT bracis-28396 43 1 moreover moreover ADV bracis-28396 43 2 , , PUNCT bracis-28396 43 3 since since SCONJ bracis-28396 43 4 features feature NOUN bracis-28396 43 5 are be AUX bracis-28396 43 6 specific specific ADJ bracis-28396 43 7 for for ADP bracis-28396 43 8 each each DET bracis-28396 43 9 problem problem NOUN bracis-28396 43 10 , , PUNCT bracis-28396 43 11 fs fs PROPN bracis-28396 43 12 has have VERB bracis-28396 43 13 to to PART bracis-28396 43 14 be be AUX bracis-28396 43 15 carried carry VERB bracis-28396 43 16 out out ADP bracis-28396 43 17 for for ADP bracis-28396 43 18 each each DET bracis-28396 43 19 data datum NOUN bracis-28396 43 20 set set VERB bracis-28396 43 21 separately separately ADV bracis-28396 43 22 . . PUNCT bracis-28396 44 1 considering consider VERB bracis-28396 44 2 that that SCONJ bracis-28396 44 3 both both CCONJ bracis-28396 44 4 the the DET bracis-28396 44 5 hp hp ADJ bracis-28396 44 6 tuning tuning NOUN bracis-28396 44 7 and and CCONJ bracis-28396 44 8 the the DET bracis-28396 44 9 task task NOUN bracis-28396 44 10 of of ADP bracis-28396 44 11 feature feature NOUN bracis-28396 44 12 selection selection NOUN bracis-28396 44 13 relies rely VERB bracis-28396 44 14 upon upon SCONJ bracis-28396 44 15 each each DET bracis-28396 44 16 other other ADJ bracis-28396 44 17 , , PUNCT bracis-28396 44 18 they they PRON bracis-28396 44 19 should should AUX bracis-28396 44 20 be be AUX bracis-28396 44 21 performed perform VERB bracis-28396 44 22 simultaneously simultaneously ADV bracis-28396 44 23 to to PART bracis-28396 44 24 generate generate VERB bracis-28396 44 25 more more ADJ bracis-28396 44 26 robust robust ADJ bracis-28396 44 27 models model NOUN bracis-28396 44 28 , , PUNCT bracis-28396 44 29 concerning concern VERB bracis-28396 44 30 generalization generalization NOUN bracis-28396 44 31 purposes purpose NOUN bracis-28396 44 32 . . PUNCT bracis-28396 45 1 in in ADP bracis-28396 45 2 general general ADJ bracis-28396 45 3 , , PUNCT bracis-28396 45 4 ml ml AUX bracis-28396 45 5 techniques technique NOUN bracis-28396 45 6 require require VERB bracis-28396 45 7 tuning tune VERB bracis-28396 45 8 more more ADJ bracis-28396 45 9 than than ADP bracis-28396 45 10 one one NUM bracis-28396 45 11 hp hp NOUN bracis-28396 45 12 , , PUNCT bracis-28396 45 13 since since SCONJ bracis-28396 45 14 the the DET bracis-28396 45 15 tackled tackle VERB bracis-28396 45 16 problems problem NOUN bracis-28396 45 17 generally generally ADV bracis-28396 45 18 are be AUX bracis-28396 45 19 described describe VERB bracis-28396 45 20 by by ADP bracis-28396 45 21 many many ADJ bracis-28396 45 22 features feature NOUN bracis-28396 45 23 , , PUNCT bracis-28396 45 24 thus thus ADV bracis-28396 45 25 implying imply VERB bracis-28396 45 26 on on ADP bracis-28396 45 27 large large ADJ bracis-28396 45 28 search search NOUN bracis-28396 45 29 spaces space NOUN bracis-28396 45 30 . . PUNCT bracis-28396 46 1 in in ADP bracis-28396 46 2 this this DET bracis-28396 46 3 context context NOUN bracis-28396 46 4 , , PUNCT bracis-28396 46 5 metaheuristics metaheuristic NOUN bracis-28396 46 6 approaches approach NOUN bracis-28396 46 7 are be AUX bracis-28396 46 8 commonly commonly ADV bracis-28396 46 9 employed employ VERB bracis-28396 46 10 to to PART bracis-28396 46 11 solve solve VERB bracis-28396 46 12 such such ADJ bracis-28396 46 13 problems problem NOUN bracis-28396 46 14 by by ADP bracis-28396 46 15 randomly randomly ADV bracis-28396 46 16 initializing initialize VERB bracis-28396 46 17 a a DET bracis-28396 46 18 collection collection NOUN bracis-28396 46 19 of of ADP bracis-28396 46 20 candidate candidate NOUN bracis-28396 46 21 solutions solution NOUN bracis-28396 46 22 , , PUNCT bracis-28396 46 23 which which PRON bracis-28396 46 24 interact interact VERB bracis-28396 46 25 among among ADP bracis-28396 46 26 themselves themselves PRON bracis-28396 46 27 and and CCONJ bracis-28396 46 28 perform perform VERB bracis-28396 46 29 a a DET bracis-28396 46 30 directed direct VERB bracis-28396 46 31 exploration exploration NOUN bracis-28396 46 32 of of ADP bracis-28396 46 33 the the DET bracis-28396 46 34 search search NOUN bracis-28396 46 35 space space NOUN bracis-28396 46 36 toward toward ADP bracis-28396 46 37 the the DET bracis-28396 46 38 results result NOUN bracis-28396 46 39 that that PRON bracis-28396 46 40 best well ADV bracis-28396 46 41 fit fit VERB bracis-28396 46 42 a a DET bracis-28396 46 43 desirable desirable ADJ bracis-28396 46 44 target target NOUN bracis-28396 46 45 function function NOUN bracis-28396 46 46 with with ADP bracis-28396 46 47 an an DET bracis-28396 46 48 acceptable acceptable ADJ bracis-28396 46 49 computational computational ADJ bracis-28396 46 50 cost cost NOUN bracis-28396 46 51 . . PUNCT bracis-28396 47 1 such such ADJ bracis-28396 47 2 approaches approach NOUN bracis-28396 47 3 are be AUX bracis-28396 47 4 commonly commonly ADV bracis-28396 47 5 employed employ VERB bracis-28396 47 6 to to PART bracis-28396 47 7 solve solve VERB bracis-28396 47 8 problems problem NOUN bracis-28396 47 9 related relate VERB bracis-28396 47 10 to to ADP bracis-28396 47 11 ml ml PROPN bracis-28396 47 12 techniques technique NOUN bracis-28396 47 13 hyperparameter hyperparameter NOUN bracis-28396 47 14 tuning tune VERB bracis-28396 47 15     SPACE bracis-28396 48 1 [ [ X bracis-28396 48 2 5 5 NUM bracis-28396 48 3 , , PUNCT bracis-28396 48 4 14 14 NUM bracis-28396 48 5 , , PUNCT bracis-28396 48 6 19 19 NUM bracis-28396 48 7 ] ] PUNCT bracis-28396 48 8 and and CCONJ bracis-28396 48 9 feature feature NOUN bracis-28396 48 10 selection selection NOUN bracis-28396 48 11     SPACE bracis-28396 49 1 [ [ X bracis-28396 49 2 15 15 NUM bracis-28396 49 3 ] ] X bracis-28396 49 4 , , PUNCT bracis-28396 49 5 among among ADP bracis-28396 49 6 others other NOUN bracis-28396 49 7     SPACE bracis-28396 50 1 [ [ X bracis-28396 50 2 4 4 NUM bracis-28396 50 3 , , PUNCT bracis-28396 50 4 22 22 NUM bracis-28396 50 5 ] ] PUNCT bracis-28396 50 6 . . PUNCT bracis-28396 51 1 in in ADP bracis-28396 51 2 this this DET bracis-28396 51 3 paper paper NOUN bracis-28396 51 4 , , PUNCT bracis-28396 51 5 we we PRON bracis-28396 51 6 investigate investigate VERB bracis-28396 51 7 the the DET bracis-28396 51 8 problem problem NOUN bracis-28396 51 9 of of ADP bracis-28396 51 10 fs f NOUN bracis-28396 51 11 and and CCONJ bracis-28396 51 12 artificial artificial ADJ bracis-28396 51 13 neural neural ADJ bracis-28396 51 14 network network NOUN bracis-28396 51 15 ( ( PUNCT bracis-28396 51 16 ann ann PROPN bracis-28396 51 17 ) ) PUNCT bracis-28396 51 18 hyperparameter hyperparameter NOUN bracis-28396 51 19 tuning tuning NOUN bracis-28396 51 20 applied apply VERB bracis-28396 51 21 in in ADP bracis-28396 51 22 the the DET bracis-28396 51 23 context context NOUN bracis-28396 51 24 of of ADP bracis-28396 51 25 wood wood NOUN bracis-28396 51 26 boards board NOUN bracis-28396 51 27 quality quality NOUN bracis-28396 51 28 classification classification NOUN bracis-28396 51 29 . . PUNCT bracis-28396 52 1 experiments experiment NOUN bracis-28396 52 2 were be AUX bracis-28396 52 3 carried carry VERB bracis-28396 52 4 out out ADP bracis-28396 52 5 using use VERB bracis-28396 52 6 the the DET bracis-28396 52 7 population population NOUN bracis-28396 52 8 - - PUNCT bracis-28396 52 9 based base VERB bracis-28396 52 10 metaheuristic metaheuristic ADJ bracis-28396 52 11 particle particle NOUN bracis-28396 52 12 swarm swarm NOUN bracis-28396 52 13 optimization optimization NOUN bracis-28396 52 14 ( ( PUNCT bracis-28396 52 15 pso pso NOUN bracis-28396 52 16 ) ) PUNCT bracis-28396 52 17     SPACE bracis-28396 53 1 [ [ X bracis-28396 53 2 9 9 NUM bracis-28396 53 3 , , PUNCT bracis-28396 53 4 20 20 NUM bracis-28396 53 5 ] ] PUNCT bracis-28396 53 6 to to PART bracis-28396 53 7 simultaneously simultaneously ADV bracis-28396 53 8 perform perform VERB bracis-28396 53 9 both both DET bracis-28396 53 10 tasks task NOUN bracis-28396 53 11 over over ADP bracis-28396 53 12 a a DET bracis-28396 53 13 multilayer multilayer ADJ bracis-28396 53 14 perceptron perceptron NOUN bracis-28396 53 15 ( ( PUNCT bracis-28396 53 16 mlp mlp PROPN bracis-28396 53 17 ) ) PUNCT bracis-28396 53 18 ann ann PROPN bracis-28396 53 19 . . PUNCT bracis-28396 54 1 moreover moreover ADV bracis-28396 54 2 , , PUNCT bracis-28396 54 3 the the DET bracis-28396 54 4 results result NOUN bracis-28396 54 5 compared compare VERB bracis-28396 54 6 against against ADP bracis-28396 54 7 five five NUM bracis-28396 54 8 distinct distinct ADJ bracis-28396 54 9 baselines baseline NOUN bracis-28396 54 10 , , PUNCT bracis-28396 54 11 as as ADV bracis-28396 54 12 well well ADV bracis-28396 54 13 as as ADP bracis-28396 54 14 a a DET bracis-28396 54 15 random random ADJ bracis-28396 54 16 search search NOUN bracis-28396 54 17 , , PUNCT bracis-28396 54 18 confirms confirm VERB bracis-28396 54 19 the the DET bracis-28396 54 20 relevance relevance NOUN bracis-28396 54 21 of of ADP bracis-28396 54 22 the the DET bracis-28396 54 23 proposed propose VERB bracis-28396 54 24 approach approach NOUN bracis-28396 54 25 . . PUNCT bracis-28396 55 1 we we PRON bracis-28396 55 2 hypothesize hypothesize VERB bracis-28396 55 3 that that SCONJ bracis-28396 55 4 the the DET bracis-28396 55 5 predictive predictive ADJ bracis-28396 55 6 performance performance NOUN bracis-28396 55 7 of of ADP bracis-28396 55 8 ann ann PROPN bracis-28396 55 9 models model NOUN bracis-28396 55 10 can can AUX bracis-28396 55 11 be be AUX bracis-28396 55 12 improved improve VERB bracis-28396 55 13 since since SCONJ bracis-28396 55 14 they they PRON bracis-28396 55 15 depend depend VERB bracis-28396 55 16 on on ADP bracis-28396 55 17 the the DET bracis-28396 55 18 hp hp ADJ bracis-28396 55 19 values value NOUN bracis-28396 55 20 and and CCONJ bracis-28396 55 21 the the DET bracis-28396 55 22 set set NOUN bracis-28396 55 23 of of ADP bracis-28396 55 24 features feature NOUN bracis-28396 55 25 used use VERB bracis-28396 55 26 to to PART bracis-28396 55 27 describe describe VERB bracis-28396 55 28 the the DET bracis-28396 55 29 problem problem NOUN bracis-28396 55 30 . . PUNCT bracis-28396 56 1 therefore therefore ADV bracis-28396 56 2 , , PUNCT bracis-28396 56 3 the the DET bracis-28396 56 4 main main ADJ bracis-28396 56 5 contributions contribution NOUN bracis-28396 56 6 of of ADP bracis-28396 56 7 this this DET bracis-28396 56 8 paper paper NOUN bracis-28396 56 9 are be AUX bracis-28396 56 10 twofold twofold ADJ bracis-28396 56 11 : : PUNCT bracis-28396 56 12 ( ( PUNCT bracis-28396 56 13 i i NOUN bracis-28396 56 14 ) ) PUNCT bracis-28396 56 15 to to PART bracis-28396 56 16 propose propose VERB bracis-28396 56 17 a a DET bracis-28396 56 18 method method NOUN bracis-28396 56 19 capable capable ADJ bracis-28396 56 20 of of ADP bracis-28396 56 21 simultaneously simultaneously ADV bracis-28396 56 22 selecting select VERB bracis-28396 56 23 the the DET bracis-28396 56 24 hyperparameters hyperparameter NOUN bracis-28396 56 25 that that PRON bracis-28396 56 26 best good ADJ bracis-28396 56 27 performs perform VERB bracis-28396 56 28 over over ADP bracis-28396 56 29 an an DET bracis-28396 56 30 mlp mlp NOUN bracis-28396 56 31 network network NOUN bracis-28396 56 32 as as ADV bracis-28396 56 33 well well ADV bracis-28396 56 34 as as ADP bracis-28396 56 35 selecting select VERB bracis-28396 56 36 the the DET bracis-28396 56 37 subset subset NOUN bracis-28396 56 38 of of ADP bracis-28396 56 39 features feature NOUN bracis-28396 56 40 that that PRON bracis-28396 56 41 best well ADV bracis-28396 56 42 describe describe VERB bracis-28396 56 43 each each DET bracis-28396 56 44 image image NOUN bracis-28396 56 45 sample sample NOUN bracis-28396 56 46 , , PUNCT bracis-28396 56 47 and and CCONJ bracis-28396 56 48 ( ( PUNCT bracis-28396 56 49 ii ii NOUN bracis-28396 56 50 ) ) PUNCT bracis-28396 56 51 to to PART bracis-28396 56 52 foster foster VERB bracis-28396 56 53 the the DET bracis-28396 56 54 scientific scientific ADJ bracis-28396 56 55 community community NOUN bracis-28396 56 56 regarding regard VERB bracis-28396 56 57 material material NOUN bracis-28396 56 58 and and CCONJ bracis-28396 56 59 wood wood NOUN bracis-28396 56 60 quality quality NOUN bracis-28396 56 61 classification classification NOUN bracis-28396 56 62 . . PUNCT bracis-28396 57 1 the the DET bracis-28396 57 2 remainder remainder NOUN bracis-28396 57 3 of of ADP bracis-28396 57 4 this this DET bracis-28396 57 5 paper paper NOUN bracis-28396 57 6 is be AUX bracis-28396 57 7 presented present VERB bracis-28396 57 8 as as SCONJ bracis-28396 57 9 follows follow VERB bracis-28396 57 10 . . PUNCT bracis-28396 58 1 section section NOUN bracis-28396 58 2     SPACE bracis-28396 58 3 2 2 NUM bracis-28396 58 4 defines define VERB bracis-28396 58 5 the the DET bracis-28396 58 6 problem problem NOUN bracis-28396 58 7 of of ADP bracis-28396 58 8 hyperparameter hyperparameter NOUN bracis-28396 58 9 tuning tuning NOUN bracis-28396 58 10 and and CCONJ bracis-28396 58 11 feature feature NOUN bracis-28396 58 12 selection selection NOUN bracis-28396 58 13 , , PUNCT bracis-28396 58 14 and and CCONJ bracis-28396 58 15 provides provide VERB bracis-28396 58 16 a a DET bracis-28396 58 17 brief brief ADJ bracis-28396 58 18 description description NOUN bracis-28396 58 19 of of ADP bracis-28396 58 20 some some DET bracis-28396 58 21 related relate VERB bracis-28396 58 22 works work NOUN bracis-28396 58 23 . . PUNCT bracis-28396 59 1 section section NOUN bracis-28396 59 2     SPACE bracis-28396 59 3 3 3 NUM bracis-28396 59 4 presents present VERB bracis-28396 59 5 the the DET bracis-28396 59 6 main main ADJ bracis-28396 59 7 concepts concept NOUN bracis-28396 59 8 of of ADP bracis-28396 59 9 ann ann PROPN bracis-28396 59 10 and and CCONJ bracis-28396 59 11 pso pso NOUN bracis-28396 59 12 . . PUNCT bracis-28396 60 1 the the DET bracis-28396 60 2 experimental experimental ADJ bracis-28396 60 3 methodology methodology NOUN bracis-28396 60 4 employed employ VERB bracis-28396 60 5 to to PART bracis-28396 60 6 evaluate evaluate VERB bracis-28396 60 7 the the DET bracis-28396 60 8 effects effect NOUN bracis-28396 60 9 of of ADP bracis-28396 60 10 fs f NOUN bracis-28396 60 11 and and CCONJ bracis-28396 60 12 mlp mlp PROPN bracis-28396 60 13 hyperparameter hyperparameter NOUN bracis-28396 60 14 tuning tune VERB bracis-28396 60 15 over over ADP bracis-28396 60 16 the the DET bracis-28396 60 17 models model NOUN bracis-28396 60 18 ’ ' PUNCT bracis-28396 60 19 performance performance NOUN bracis-28396 60 20 is be AUX bracis-28396 60 21 described describe VERB bracis-28396 60 22 in in ADP bracis-28396 60 23 sect sect NOUN bracis-28396 60 24 . . PUNCT bracis-28396 60 25     SPACE bracis-28396 61 1 4 4 X bracis-28396 61 2 . . PUNCT bracis-28396 61 3 results result NOUN bracis-28396 61 4 are be AUX bracis-28396 61 5 presented present VERB bracis-28396 61 6 and and CCONJ bracis-28396 61 7 discussed discuss VERB bracis-28396 61 8 in in ADP bracis-28396 61 9 sect sect NOUN bracis-28396 61 10 . . PUNCT bracis-28396 61 11     SPACE bracis-28396 61 12 5 5 NUM bracis-28396 61 13 , , PUNCT bracis-28396 61 14 and and CCONJ bracis-28396 61 15 finally finally ADV bracis-28396 61 16 conclusions conclusion NOUN bracis-28396 61 17 are be AUX bracis-28396 61 18 presented present VERB bracis-28396 61 19 in in ADP bracis-28396 61 20 sect sect NOUN bracis-28396 61 21 . . PUNCT bracis-28396 61 22     SPACE bracis-28396 62 1 6 6 NUM bracis-28396 62 2 . . SYM bracis-28396 62 3 2 2 NUM bracis-28396 62 4 problem problem NOUN bracis-28396 62 5 definition definition NOUN bracis-28396 62 6 and and CCONJ bracis-28396 62 7     SPACE bracis-28396 62 8 related relate VERB bracis-28396 62 9 work work NOUN bracis-28396 62 10 hyperparameter hyperparameter NOUN bracis-28396 62 11 tuning tuning NOUN bracis-28396 62 12 and and CCONJ bracis-28396 62 13 feature feature NOUN bracis-28396 62 14 subset subset NOUN bracis-28396 62 15 selection selection NOUN bracis-28396 62 16 are be AUX bracis-28396 62 17 two two NUM bracis-28396 62 18 widely widely ADV bracis-28396 62 19 employed employ VERB bracis-28396 62 20 tasks task NOUN bracis-28396 62 21 carried carry VERB bracis-28396 62 22 out out ADP bracis-28396 62 23 in in ADP bracis-28396 62 24 the the DET bracis-28396 62 25 data data NOUN bracis-28396 62 26 mining mining NOUN bracis-28396 62 27 context context NOUN bracis-28396 62 28 , , PUNCT bracis-28396 62 29 aiming aim VERB bracis-28396 62 30 to to PART bracis-28396 62 31 improve improve VERB bracis-28396 62 32 models model NOUN bracis-28396 62 33 ’ ' PUNCT bracis-28396 62 34 predictive predictive ADJ bracis-28396 62 35 performance performance NOUN bracis-28396 62 36 as as ADV bracis-28396 62 37 well well ADV bracis-28396 62 38 as as ADP bracis-28396 62 39 simplifying simplify VERB bracis-28396 62 40 them they PRON bracis-28396 62 41 . . PUNCT bracis-28396 63 1 therefore therefore ADV bracis-28396 63 2 , , PUNCT bracis-28396 63 3 this this DET bracis-28396 63 4 section section NOUN bracis-28396 63 5 formalizes formalize VERB bracis-28396 63 6 the the DET bracis-28396 63 7 problem problem NOUN bracis-28396 63 8 of of ADP bracis-28396 63 9 simultaneously simultaneously ADV bracis-28396 63 10 performing perform VERB bracis-28396 63 11 these these DET bracis-28396 63 12 two two NUM bracis-28396 63 13 tasks task NOUN bracis-28396 63 14 . . PUNCT bracis-28396 64 1 besides besides SCONJ bracis-28396 64 2 , , PUNCT bracis-28396 64 3 it it PRON bracis-28396 64 4 presents present VERB bracis-28396 64 5 an an DET bracis-28396 64 6 overview overview NOUN bracis-28396 64 7 of of ADP bracis-28396 64 8 studies study NOUN bracis-28396 64 9 related relate VERB bracis-28396 64 10 to to ADP bracis-28396 64 11 wood wood NOUN bracis-28396 64 12 quality quality NOUN bracis-28396 64 13 classification classification NOUN bracis-28396 64 14 and and CCONJ bracis-28396 64 15 the the DET bracis-28396 64 16 importance importance NOUN bracis-28396 64 17 of of ADP bracis-28396 64 18 hp hp ADJ bracis-28396 64 19 tuning tuning NOUN bracis-28396 64 20 and and CCONJ bracis-28396 64 21 fs fs PROPN bracis-28396 64 22 . . PROPN bracis-28396 64 23 2.1 2.1 NUM bracis-28396 64 24 problem problem NOUN bracis-28396 64 25 definition definition NOUN bracis-28396 64 26 the the DET bracis-28396 64 27 problem problem NOUN bracis-28396 64 28 investigated investigate VERB bracis-28396 64 29 in in ADP bracis-28396 64 30 this this DET bracis-28396 64 31 paper paper NOUN bracis-28396 64 32 consists consist VERB bracis-28396 64 33 of of ADP bracis-28396 64 34 tuning tune VERB bracis-28396 64 35 the the DET bracis-28396 64 36 hp hp ADJ bracis-28396 64 37 values value NOUN bracis-28396 64 38 of of ADP bracis-28396 64 39 an an DET bracis-28396 64 40 mlp mlp NOUN bracis-28396 64 41 artificial artificial ADJ bracis-28396 64 42 neural neural ADJ bracis-28396 64 43 network network NOUN bracis-28396 64 44 algorithm algorithm NOUN bracis-28396 64 45 , , PUNCT bracis-28396 64 46 as as ADV bracis-28396 64 47 well well ADV bracis-28396 64 48 as as ADP bracis-28396 64 49 selecting select VERB bracis-28396 64 50 a a DET bracis-28396 64 51 subset subset NOUN bracis-28396 64 52 of of ADP bracis-28396 64 53 features feature NOUN bracis-28396 64 54 that that PRON bracis-28396 64 55 are be AUX bracis-28396 64 56 relevant relevant ADJ bracis-28396 64 57 for for ADP bracis-28396 64 58 the the DET bracis-28396 64 59 problem problem NOUN bracis-28396 64 60 of of ADP bracis-28396 64 61 wood wood NOUN bracis-28396 64 62 quality quality NOUN bracis-28396 64 63 classification classification NOUN bracis-28396 64 64 towards towards ADP bracis-28396 64 65 the the DET bracis-28396 64 66 improvement improvement NOUN bracis-28396 64 67 of of ADP bracis-28396 64 68 the the DET bracis-28396 64 69 models model NOUN bracis-28396 64 70 ’ ' PUNCT bracis-28396 64 71 predictive predictive ADJ bracis-28396 64 72 performance performance NOUN bracis-28396 64 73 . . PUNCT bracis-28396 65 1 let let VERB bracis-28396 65 2 a a DET bracis-28396 65 3 be be AUX bracis-28396 65 4 an an DET bracis-28396 65 5 mlp mlp NOUN bracis-28396 65 6 algorithm algorithm NOUN bracis-28396 65 7 that that PRON bracis-28396 65 8 comprises comprise VERB bracis-28396 65 9 the the DET bracis-28396 65 10 hyperparameter hyperparameter NOUN bracis-28396 65 11 space space NOUN bracis-28396 65 12 \(\varlambda \(\varlambda NOUN bracis-28396 65 13 \ \ PROPN bracis-28396 65 14 ) ) PUNCT bracis-28396 65 15 . . PUNCT bracis-28396 66 1 for for ADP bracis-28396 66 2 each each DET bracis-28396 66 3 hyperparameter hyperparameter NOUN bracis-28396 66 4 setting set VERB bracis-28396 66 5 \(\lambda \(\lambda NOUN bracis-28396 66 6 \in \in ADJ bracis-28396 66 7 \varlambda \varlambda PROPN bracis-28396 66 8 \ \ NOUN bracis-28396 66 9 ) ) PUNCT bracis-28396 66 10 , , PUNCT bracis-28396 66 11 let let VERB bracis-28396 66 12 \(a_{\lambda \(a_{\lambda NOUN bracis-28396 66 13 } } PUNCT bracis-28396 66 14 \ \ NOUN bracis-28396 66 15 ) ) PUNCT bracis-28396 66 16 represent represent VERB bracis-28396 66 17 the the DET bracis-28396 66 18 learning learning NOUN bracis-28396 66 19 algorithm algorithm NOUN bracis-28396 66 20 a a PRON bracis-28396 66 21 that that PRON bracis-28396 66 22 employs employ VERB bracis-28396 66 23 the the DET bracis-28396 66 24 hyperparameter hyperparameter NOUN bracis-28396 66 25 setting set VERB bracis-28396 66 26 \(\lambda \(\lambda NOUN bracis-28396 66 27 \ \ NOUN bracis-28396 66 28 ) ) PUNCT bracis-28396 66 29 . . PUNCT bracis-28396 67 1 also also ADV bracis-28396 67 2 , , PUNCT bracis-28396 67 3 consider consider VERB bracis-28396 67 4 \(d \(d NOUN bracis-28396 67 5 = = SYM bracis-28396 67 6 \{(x_1,y_1 \{(x_1,y_1 PROPN bracis-28396 67 7 ) ) PUNCT bracis-28396 67 8 , , PUNCT bracis-28396 67 9 ( ( PUNCT bracis-28396 67 10 x_2,y_2 x_2,y_2 PROPN bracis-28396 67 11 ) ) PUNCT bracis-28396 67 12 , , PUNCT bracis-28396 67 13 \ldots \ldots PROPN bracis-28396 67 14 , , PUNCT bracis-28396 67 15 ( ( PUNCT bracis-28396 67 16 x_n x_n PROPN bracis-28396 67 17 , , PUNCT bracis-28396 67 18 y_n)\}\ y_n)\}\ NOUN bracis-28396 67 19 ) ) PUNCT bracis-28396 67 20 a a DET bracis-28396 67 21 dataset dataset NOUN bracis-28396 67 22 composed compose VERB bracis-28396 67 23 of of ADP bracis-28396 67 24 n n DET bracis-28396 67 25 instances instance NOUN bracis-28396 67 26 , , PUNCT bracis-28396 67 27 such such ADJ bracis-28396 67 28 that that SCONJ bracis-28396 67 29 \(\boldsymbol{x \(\boldsymbol{x ADJ bracis-28396 67 30 } } PUNCT bracis-28396 67 31 \in \in ADJ bracis-28396 67 32 \mathbb \mathbb PROPN bracis-28396 67 33 { { PUNCT bracis-28396 67 34 r}^{m}\ r}^{m}\ NOUN bracis-28396 67 35 ) ) PUNCT bracis-28396 67 36 is be AUX bracis-28396 67 37 a a DET bracis-28396 67 38 feature feature NOUN bracis-28396 67 39 vector vector NOUN bracis-28396 67 40 and and CCONJ bracis-28396 67 41 y y NOUN bracis-28396 67 42 is be AUX bracis-28396 67 43 the the DET bracis-28396 67 44 target target NOUN bracis-28396 67 45 value value NOUN bracis-28396 67 46 . . PUNCT bracis-28396 68 1 moreover moreover ADV bracis-28396 68 2 , , PUNCT bracis-28396 68 3 one one PRON bracis-28396 68 4 can can AUX bracis-28396 68 5 define define VERB bracis-28396 68 6 \(\kappa \(\kappa PROPN bracis-28396 68 7 \ \ NOUN bracis-28396 68 8 ) ) PUNCT bracis-28396 68 9 as as ADP bracis-28396 68 10 the the DET bracis-28396 68 11 subset subset NOUN bracis-28396 68 12 of of ADP bracis-28396 68 13 feature feature NOUN bracis-28396 68 14 from from ADP bracis-28396 68 15 \(\boldsymbol{x}\ \(\boldsymbol{x}\ PROPN bracis-28396 68 16 ) ) PUNCT bracis-28396 68 17 . . PUNCT bracis-28396 69 1 finally finally ADV bracis-28396 69 2 , , PUNCT bracis-28396 69 3 let let VERB bracis-28396 69 4 \(a^{\kappa \(a^{\kappa PROPN bracis-28396 69 5 } } PUNCT bracis-28396 69 6 \ \ NOUN bracis-28396 69 7 ) ) PUNCT bracis-28396 69 8 be be AUX bracis-28396 69 9 an an DET bracis-28396 69 10 algorithm algorithm NOUN bracis-28396 69 11 a a DET bracis-28396 69 12 trained train VERB bracis-28396 69 13 with with ADP bracis-28396 69 14 a a DET bracis-28396 69 15 subset subset NOUN bracis-28396 69 16 of of ADP bracis-28396 69 17 features feature NOUN bracis-28396 69 18 \(\kappa \(\kappa PROPN bracis-28396 69 19 \ \ NOUN bracis-28396 69 20 ) ) PUNCT bracis-28396 69 21     SPACE bracis-28396 70 1 [ [ X bracis-28396 70 2 12 12 NUM bracis-28396 70 3 ] ] PUNCT bracis-28396 70 4 . . PUNCT bracis-28396 71 1 therefore therefore ADV bracis-28396 71 2 , , PUNCT bracis-28396 71 3 the the DET bracis-28396 71 4 main main ADJ bracis-28396 71 5 goal goal NOUN bracis-28396 71 6 of of ADP bracis-28396 71 7 hp hp ADJ bracis-28396 71 8 tuning tuning NOUN bracis-28396 71 9 is be AUX bracis-28396 71 10 to to ADP bracis-28396 71 11 finding find VERB bracis-28396 71 12 \(\lambda \(\lambda NOUN bracis-28396 71 13 ^ ^ NOUN bracis-28396 71 14 * * PUNCT bracis-28396 71 15 = = SYM bracis-28396 71 16 \text \text ADJ bracis-28396 71 17 { { PUNCT bracis-28396 71 18 arg arg NOUN bracis-28396 71 19 min}_{\lambda min}_{\lambda X bracis-28396 71 20 \in \in ADJ bracis-28396 71 21 \varlambda \varlambda PRON bracis-28396 71 22 } } PUNCT bracis-28396 71 23 m(a_\lambda m(a_\lambda NOUN bracis-28396 71 24 , , PUNCT bracis-28396 71 25 d d X bracis-28396 71 26 ) ) PUNCT bracis-28396 71 27 \in \in VERB bracis-28396 71 28 \varlambda \varlambda NOUN bracis-28396 71 29 \ \ NOUN bracis-28396 71 30 ) ) PUNCT bracis-28396 71 31 that that PRON bracis-28396 71 32 minimizes minimize VERB bracis-28396 71 33 some some DET bracis-28396 71 34 loss loss NOUN bracis-28396 71 35 function function NOUN bracis-28396 71 36 , , PUNCT bracis-28396 71 37 such such ADJ bracis-28396 71 38 as as ADP bracis-28396 71 39 the the DET bracis-28396 71 40 misclassification misclassification NOUN bracis-28396 71 41 rate rate NOUN bracis-28396 71 42 using use VERB bracis-28396 71 43 the the DET bracis-28396 71 44 algorithm algorithm NOUN bracis-28396 71 45 a a PRON bracis-28396 71 46 over over ADP bracis-28396 71 47 instances instance NOUN bracis-28396 71 48 not not PART bracis-28396 71 49 used use VERB bracis-28396 71 50 for for ADP bracis-28396 71 51 training training NOUN bracis-28396 71 52 purposes purpose NOUN bracis-28396 71 53 . . PUNCT bracis-28396 72 1 moreover moreover ADV bracis-28396 72 2 , , PUNCT bracis-28396 72 3 one one PRON bracis-28396 72 4 can can AUX bracis-28396 72 5 estimate estimate VERB bracis-28396 72 6 the the DET bracis-28396 72 7 misclassification misclassification NOUN bracis-28396 72 8 rate rate NOUN bracis-28396 72 9 m(a m(a NOUN bracis-28396 72 10 , , PUNCT bracis-28396 72 11     SPACE bracis-28396 72 12 d d NOUN bracis-28396 72 13 ) ) PUNCT bracis-28396 72 14 achieved achieve VERB bracis-28396 72 15 by by ADP bracis-28396 72 16 a a DET bracis-28396 72 17 when when SCONJ bracis-28396 72 18 trained train VERB bracis-28396 72 19 and and CCONJ bracis-28396 72 20 tested test VERB bracis-28396 72 21 on on ADP bracis-28396 72 22 d d PROPN bracis-28396 72 23 through through ADP bracis-28396 72 24 a a DET bracis-28396 72 25 stratified stratified ADJ bracis-28396 72 26 multi multi ADJ bracis-28396 72 27 - - ADJ bracis-28396 72 28 fold fold ADJ bracis-28396 72 29 cross cross ADJ bracis-28396 72 30 - - ADJ bracis-28396 72 31 validation validation ADJ bracis-28396 72 32 resampling resampling NOUN bracis-28396 72 33 method method NOUN bracis-28396 72 34 . . PUNCT bracis-28396 73 1 similarly similarly ADV bracis-28396 73 2 to to ADP bracis-28396 73 3 hp hp PROPN bracis-28396 73 4 tuning tuning NOUN bracis-28396 73 5 , , PUNCT bracis-28396 73 6 the the DET bracis-28396 73 7 goal goal NOUN bracis-28396 73 8 of of ADP bracis-28396 73 9 feature feature NOUN bracis-28396 73 10 subset subset NOUN bracis-28396 73 11 selection selection NOUN bracis-28396 73 12 is be AUX bracis-28396 73 13 to to PART bracis-28396 73 14 find find VERB bracis-28396 73 15 \(\kappa \(\kappa PRON bracis-28396 73 16 ^*=\text ^*=\text X bracis-28396 73 17 { { PUNCT bracis-28396 73 18 arg arg PROPN bracis-28396 73 19 min}_{\kappa min}_{\kappa PROPN bracis-28396 73 20 \subseteq \subseteq PROPN bracis-28396 73 21 \boldsymbol{x}}m(a^\kappa \boldsymbol{x}}m(a^\kappa NOUN bracis-28396 73 22 , , PUNCT bracis-28396 73 23 d d X bracis-28396 73 24 ) ) PUNCT bracis-28396 73 25 \subseteq \subseteq PROPN bracis-28396 73 26 \boldsymbol{x}\ \boldsymbol{x}\ NUM bracis-28396 73 27 ) ) PUNCT bracis-28396 73 28 which which PRON bracis-28396 73 29 minimizes minimize VERB bracis-28396 73 30 the the DET bracis-28396 73 31 loss loss NOUN bracis-28396 73 32 function function NOUN bracis-28396 73 33 achieved achieve VERB bracis-28396 73 34 by by ADP bracis-28396 73 35 a a PRON bracis-28396 73 36 when when SCONJ bracis-28396 73 37 trained train VERB bracis-28396 73 38 on on ADP bracis-28396 73 39 d. d. PROPN bracis-28396 73 40 therefore therefore ADV bracis-28396 73 41 , , PUNCT bracis-28396 73 42 the the DET bracis-28396 73 43 aim aim NOUN bracis-28396 73 44 of of ADP bracis-28396 73 45 combining combine VERB bracis-28396 73 46 hyperparameter hyperparameter NOUN bracis-28396 73 47 tuning tuning NOUN bracis-28396 73 48 and and CCONJ bracis-28396 73 49 feature feature NOUN bracis-28396 73 50 subset subset NOUN bracis-28396 73 51 selection selection NOUN bracis-28396 73 52 is be AUX bracis-28396 73 53 to to PART bracis-28396 73 54 find find VERB bracis-28396 73 55 the the DET bracis-28396 73 56 hyperparameter hyperparameter NOUN bracis-28396 73 57 setting set VERB bracis-28396 73 58 \(\lambda \(\lambda NOUN bracis-28396 73 59 ^ ^ NOUN bracis-28396 73 60 * * PUNCT bracis-28396 73 61 \in \in ADJ bracis-28396 73 62 \varlambda \varlambda PROPN bracis-28396 73 63 \ \ NOUN bracis-28396 73 64 ) ) PUNCT bracis-28396 73 65 and and CCONJ bracis-28396 73 66 the the DET bracis-28396 73 67 feature feature NOUN bracis-28396 73 68 subset subset VERB bracis-28396 73 69 \(\kappa \(\kappa NOUN bracis-28396 73 70 ^ ^ PUNCT bracis-28396 73 71 * * PROPN bracis-28396 73 72 \subseteq \subseteq PROPN bracis-28396 73 73 \boldsymbol{x}\ \boldsymbol{x}\ NUM bracis-28396 73 74 ) ) PUNCT bracis-28396 73 75 which which PRON bracis-28396 73 76 has have VERB bracis-28396 73 77 the the DET bracis-28396 73 78 lowest low ADJ bracis-28396 73 79 misclassification misclassification NOUN bracis-28396 73 80 rate rate NOUN bracis-28396 73 81 among among ADP bracis-28396 73 82 all all DET bracis-28396 73 83 hp hp ADJ bracis-28396 73 84 settings setting NOUN bracis-28396 73 85 and and CCONJ bracis-28396 73 86 features feature VERB bracis-28396 73 87 subsets subset NOUN bracis-28396 73 88 , , PUNCT bracis-28396 73 89 i.e. i.e. X bracis-28396 73 90 , , PUNCT bracis-28396 73 91 \((\kappa \((\kappa NOUN bracis-28396 73 92 ^*,\lambda ^*,\lambda PRON bracis-28396 73 93 ^ ^ NOUN bracis-28396 73 94 * * PUNCT bracis-28396 73 95 ) ) PUNCT bracis-28396 74 1 = = SYM bracis-28396 74 2 \text \text ADJ bracis-28396 74 3 { { PUNCT bracis-28396 74 4 arg arg NOUN bracis-28396 74 5 min}_{\lambda min}_{\lambda X bracis-28396 74 6 \in \in PROPN bracis-28396 74 7 \varlambda \varlambda NOUN bracis-28396 74 8 , , PUNCT bracis-28396 74 9 \kappa \kappa VERB bracis-28396 74 10 \subseteq \subseteq INTJ bracis-28396 74 11 \boldsymbol{x \boldsymbol{x ADP bracis-28396 74 12 } } PUNCT bracis-28396 74 13 } } PUNCT bracis-28396 74 14 m(a^{\kappa m(a^{\kappa ADV bracis-28396 74 15 } } PUNCT bracis-28396 74 16 _ _ PUNCT bracis-28396 74 17 { { PUNCT bracis-28396 74 18 \lambda \lambda PROPN bracis-28396 74 19 } } PUNCT bracis-28396 74 20 , , PUNCT bracis-28396 74 21 d)\ d)\ ADV bracis-28396 74 22 ) ) PUNCT bracis-28396 74 23 . . PUNCT bracis-28396 75 1 2.2 2.2 NUM bracis-28396 75 2 related relate VERB bracis-28396 75 3 work work NOUN bracis-28396 75 4 one one NUM bracis-28396 75 5 of of ADP bracis-28396 75 6 the the DET bracis-28396 75 7 first first ADJ bracis-28396 75 8 studies study NOUN bracis-28396 75 9 to to PART bracis-28396 75 10 investigate investigate VERB bracis-28396 75 11 the the DET bracis-28396 75 12 problem problem NOUN bracis-28396 75 13 of of ADP bracis-28396 75 14 wood wood NOUN bracis-28396 75 15 quality quality NOUN bracis-28396 75 16 control control NOUN bracis-28396 75 17 using use VERB bracis-28396 75 18 an an DET bracis-28396 75 19 automated automate VERB bracis-28396 75 20 visual visual ADJ bracis-28396 75 21 inspection inspection NOUN bracis-28396 75 22 system system NOUN bracis-28396 75 23 based base VERB bracis-28396 75 24 on on ADP bracis-28396 75 25 machine machine NOUN bracis-28396 75 26 learning learning NOUN bracis-28396 75 27 techniques technique NOUN bracis-28396 75 28 was be AUX bracis-28396 75 29 accomplished accomplish VERB bracis-28396 75 30 by by ADP bracis-28396 75 31     SPACE bracis-28396 75 32 [ [ X bracis-28396 75 33 16 16 NUM bracis-28396 75 34 ] ] PUNCT bracis-28396 75 35 . . PUNCT bracis-28396 76 1 since since SCONJ bracis-28396 76 2 then then ADV bracis-28396 76 3 , , PUNCT bracis-28396 76 4 many many ADJ bracis-28396 76 5 others other NOUN bracis-28396 76 6 have have AUX bracis-28396 76 7 investigated investigate VERB bracis-28396 76 8 this this DET bracis-28396 76 9 problem problem NOUN bracis-28396 76 10 aiming aim VERB bracis-28396 76 11 to to PART bracis-28396 76 12 improve improve VERB bracis-28396 76 13 predictive predictive ADJ bracis-28396 76 14 performance performance NOUN bracis-28396 76 15 by by ADP bracis-28396 76 16 analyzing analyze VERB bracis-28396 76 17 and and CCONJ bracis-28396 76 18 selecting select VERB bracis-28396 76 19 different different ADJ bracis-28396 76 20 features feature NOUN bracis-28396 76 21 , , PUNCT bracis-28396 76 22 as as ADV bracis-28396 76 23 well well ADV bracis-28396 76 24 as as ADP bracis-28396 76 25 optimizing optimize VERB bracis-28396 76 26 hp hp ADJ bracis-28396 76 27 values value NOUN bracis-28396 76 28 of of ADP bracis-28396 76 29 ml ml NOUN bracis-28396 76 30 techniques technique NOUN bracis-28396 76 31     SPACE bracis-28396 77 1 [ [ X bracis-28396 77 2 21 21 NUM bracis-28396 77 3 , , PUNCT bracis-28396 77 4 24 24 NUM bracis-28396 77 5 ] ] PUNCT bracis-28396 77 6 . . PUNCT bracis-28396 78 1 tiryaki tiryaki PROPN bracis-28396 78 2 et et PROPN bracis-28396 78 3 al al PROPN bracis-28396 78 4 . . PUNCT bracis-28396 78 5     SPACE bracis-28396 79 1 [ [ X bracis-28396 79 2 24 24 NUM bracis-28396 79 3 ] ] PUNCT bracis-28396 79 4 employed employ VERB bracis-28396 79 5 an an DET bracis-28396 79 6 ann ann NOUN bracis-28396 79 7 for for ADP bracis-28396 79 8 modeling model VERB bracis-28396 79 9 the the DET bracis-28396 79 10 wood wood NOUN bracis-28396 79 11 surface surface NOUN bracis-28396 79 12 roughness roughness NOUN bracis-28396 79 13 in in ADP bracis-28396 79 14 the the DET bracis-28396 79 15 machining machining NOUN bracis-28396 79 16 process process NOUN bracis-28396 79 17 . . PUNCT bracis-28396 80 1 the the DET bracis-28396 80 2 study study NOUN bracis-28396 80 3 highlights highlight VERB bracis-28396 80 4 some some DET bracis-28396 80 5 variables variable NOUN bracis-28396 80 6 that that PRON bracis-28396 80 7 have have VERB bracis-28396 80 8 impact impact NOUN bracis-28396 80 9 and and CCONJ bracis-28396 80 10 influence influence NOUN bracis-28396 80 11 in in ADP bracis-28396 80 12 the the DET bracis-28396 80 13 surface surface NOUN bracis-28396 80 14 roughness roughness NOUN bracis-28396 80 15 , , PUNCT bracis-28396 80 16 such such ADJ bracis-28396 80 17 as as ADP bracis-28396 80 18 wood wood NOUN bracis-28396 80 19 species specie NOUN bracis-28396 80 20 , , PUNCT bracis-28396 80 21 the the DET bracis-28396 80 22 feed feed NOUN bracis-28396 80 23 rate rate NOUN bracis-28396 80 24 , , PUNCT bracis-28396 80 25 the the DET bracis-28396 80 26 number number NOUN bracis-28396 80 27 of of ADP bracis-28396 80 28 the the DET bracis-28396 80 29 cutter cutter NOUN bracis-28396 80 30 , , PUNCT bracis-28396 80 31 and and CCONJ bracis-28396 80 32 the the DET bracis-28396 80 33 cutting cut VERB bracis-28396 80 34 depth depth NOUN bracis-28396 80 35 . . PUNCT bracis-28396 81 1 the the DET bracis-28396 81 2 model model NOUN bracis-28396 81 3 ’s ’s PART bracis-28396 81 4 predictive predictive ADJ bracis-28396 81 5 performance performance NOUN bracis-28396 81 6 was be AUX bracis-28396 81 7 good good ADJ bracis-28396 81 8 enough enough ADV bracis-28396 81 9 to to PART bracis-28396 81 10 allow allow VERB bracis-28396 81 11 its its PRON bracis-28396 81 12 application application NOUN bracis-28396 81 13 in in ADP bracis-28396 81 14 the the DET bracis-28396 81 15 wood wood NOUN bracis-28396 81 16 industry industry NOUN bracis-28396 81 17 in in ADP bracis-28396 81 18 order order NOUN bracis-28396 81 19 to to PART bracis-28396 81 20 optimize optimize VERB bracis-28396 81 21 effort effort NOUN bracis-28396 81 22 , , PUNCT bracis-28396 81 23 time time NOUN bracis-28396 81 24 , , PUNCT bracis-28396 81 25 and and CCONJ bracis-28396 81 26 energy energy NOUN bracis-28396 81 27 . . PUNCT bracis-28396 82 1 others other NOUN bracis-28396 82 2 addressed address VERB bracis-28396 82 3 the the DET bracis-28396 82 4 problem problem NOUN bracis-28396 82 5 of of ADP bracis-28396 82 6 combining combine VERB bracis-28396 82 7 fs fs ADP bracis-28396 82 8 and and CCONJ bracis-28396 82 9 hp hp ADJ bracis-28396 82 10 tuning tune VERB bracis-28396 82 11 for for ADP bracis-28396 82 12 ml ml NOUN bracis-28396 82 13 techniques technique NOUN bracis-28396 82 14 over over ADP bracis-28396 82 15 different different ADJ bracis-28396 82 16 applications application NOUN bracis-28396 82 17 . . PUNCT bracis-28396 83 1 as as SCONJ bracis-28396 83 2 previously previously ADV bracis-28396 83 3 mentioned mention VERB bracis-28396 83 4 , , PUNCT bracis-28396 83 5 some some DET bracis-28396 83 6 techniques technique NOUN bracis-28396 83 7 are be AUX bracis-28396 83 8 more more ADV bracis-28396 83 9 sensitive sensitive ADJ bracis-28396 83 10 to to ADP bracis-28396 83 11 hp hp ADJ bracis-28396 83 12 tuning tuning NOUN bracis-28396 83 13 and and CCONJ bracis-28396 83 14 fs f NOUN bracis-28396 83 15 than than ADP bracis-28396 83 16 others other NOUN bracis-28396 83 17 . . PUNCT bracis-28396 84 1 besides besides SCONJ bracis-28396 84 2 , , PUNCT bracis-28396 84 3 some some DET bracis-28396 84 4 optimization optimization NOUN bracis-28396 84 5 methods method NOUN bracis-28396 84 6 , , PUNCT bracis-28396 84 7 such such ADJ bracis-28396 84 8 as as ADP bracis-28396 84 9 metaheuristic metaheuristic ADJ bracis-28396 84 10 approaches approach NOUN bracis-28396 84 11 based base VERB bracis-28396 84 12 on on ADP bracis-28396 84 13 evolutionary evolutionary ADJ bracis-28396 84 14 algorithms algorithm NOUN bracis-28396 84 15 and and CCONJ bracis-28396 84 16 swarm swarm NOUN bracis-28396 84 17 intelligence intelligence NOUN bracis-28396 84 18     SPACE bracis-28396 85 1 [ [ X bracis-28396 85 2 10 10 NUM bracis-28396 85 3 ] ] PUNCT bracis-28396 85 4 , , PUNCT bracis-28396 85 5 for for ADP bracis-28396 85 6 instance instance NOUN bracis-28396 85 7 , , PUNCT bracis-28396 85 8 have have AUX bracis-28396 85 9 successfully successfully ADV bracis-28396 85 10 accomplished accomplish VERB bracis-28396 85 11 the the DET bracis-28396 85 12 task task NOUN bracis-28396 85 13 . . PUNCT bracis-28396 86 1 in in ADP bracis-28396 86 2 this this DET bracis-28396 86 3 context context NOUN bracis-28396 86 4 , , PUNCT bracis-28396 86 5 roder roder PROPN bracis-28396 86 6 et et PROPN bracis-28396 86 7 al al PROPN bracis-28396 86 8 . . PUNCT bracis-28396 86 9     SPACE bracis-28396 87 1 [ [ X bracis-28396 87 2 21 21 NUM bracis-28396 87 3 ] ] PUNCT bracis-28396 87 4 used use VERB bracis-28396 87 5 the the DET bracis-28396 87 6 pso pso NOUN bracis-28396 87 7 algorithm algorithm NOUN bracis-28396 87 8 to to PART bracis-28396 87 9 tune tune VERB bracis-28396 87 10 the the DET bracis-28396 87 11 hp hp NOUN bracis-28396 87 12 of of ADP bracis-28396 87 13 an an DET bracis-28396 87 14 ann ann PROPN bracis-28396 87 15 applied apply VERB bracis-28396 87 16 for for ADP bracis-28396 87 17 wood wood NOUN bracis-28396 87 18 quality quality NOUN bracis-28396 87 19 classification classification NOUN bracis-28396 87 20 aiming aim VERB bracis-28396 87 21 to to PART bracis-28396 87 22 enhance enhance VERB bracis-28396 87 23 the the DET bracis-28396 87 24 model model NOUN bracis-28396 87 25 ’s ’s PART bracis-28396 87 26 predictive predictive ADJ bracis-28396 87 27 performance performance NOUN bracis-28396 87 28 . . PUNCT bracis-28396 88 1 the the DET bracis-28396 88 2 authors author NOUN bracis-28396 88 3 employed employ VERB bracis-28396 88 4 the the DET bracis-28396 88 5 glcm glcm NOUN bracis-28396 88 6     SPACE bracis-28396 88 7 [ [ X bracis-28396 88 8 8 8 NUM bracis-28396 88 9 ] ] PUNCT bracis-28396 88 10 to to PART bracis-28396 88 11 extract extract VERB bracis-28396 88 12 features feature NOUN bracis-28396 88 13 from from ADP bracis-28396 88 14 the the DET bracis-28396 88 15 same same ADJ bracis-28396 88 16 dataset dataset NOUN bracis-28396 88 17 used use VERB bracis-28396 88 18 by by ADP bracis-28396 88 19     SPACE bracis-28396 88 20 [ [ X bracis-28396 88 21 2 2 NUM bracis-28396 88 22 ] ] PUNCT bracis-28396 88 23 , , PUNCT bracis-28396 88 24 which which PRON bracis-28396 88 25 is be AUX bracis-28396 88 26 composed compose VERB bracis-28396 88 27 of of ADP bracis-28396 88 28 five five NUM bracis-28396 88 29 statistical statistical ADJ bracis-28396 88 30 measures measure NOUN bracis-28396 88 31 for for ADP bracis-28396 88 32 two two NUM bracis-28396 88 33 angles angle NOUN bracis-28396 88 34 , , PUNCT bracis-28396 88 35 i.e. i.e. X bracis-28396 88 36 , , PUNCT bracis-28396 88 37 0 0 NUM bracis-28396 88 38 and and CCONJ bracis-28396 88 39 \(90^\circ \(90^\circ ADJ bracis-28396 88 40 \ \ PROPN bracis-28396 88 41 ): ): PUNCT bracis-28396 88 42 entropy entropy PROPN bracis-28396 88 43 , , PUNCT bracis-28396 88 44 energy energy NOUN bracis-28396 88 45 , , PUNCT bracis-28396 88 46 maximum maximum ADJ bracis-28396 88 47 intensity intensity NOUN bracis-28396 88 48 , , PUNCT bracis-28396 88 49 inverse inverse ADJ bracis-28396 88 50 difference difference NOUN bracis-28396 88 51 moment moment NOUN bracis-28396 88 52 , , PUNCT bracis-28396 88 53 and and CCONJ bracis-28396 88 54 correlation correlation NOUN bracis-28396 88 55 . . PUNCT bracis-28396 89 1 experimental experimental ADJ bracis-28396 89 2 results result NOUN bracis-28396 89 3 obtained obtain VERB bracis-28396 89 4 up up ADP bracis-28396 89 5 to to ADP bracis-28396 89 6 \(6\%\ \(6\%\ NOUN bracis-28396 89 7 ) ) PUNCT bracis-28396 89 8 of of ADP bracis-28396 89 9 accuracy accuracy NOUN bracis-28396 89 10 gain gain NOUN bracis-28396 89 11 concerning concern VERB bracis-28396 89 12 the the DET bracis-28396 89 13 ann ann PROPN bracis-28396 89 14 classification classification NOUN bracis-28396 89 15 , , PUNCT bracis-28396 89 16 and and CCONJ bracis-28396 89 17 corroborate corroborate VERB bracis-28396 89 18 the the DET bracis-28396 89 19 necessity necessity NOUN bracis-28396 89 20 of of ADP bracis-28396 89 21 tuning tune VERB bracis-28396 89 22 the the DET bracis-28396 89 23 ann ann PROPN bracis-28396 89 24 hyperparameters hyperparameter NOUN bracis-28396 89 25 ; ; PUNCT bracis-28396 89 26 stating state VERB bracis-28396 89 27 the the DET bracis-28396 89 28 efficiency efficiency NOUN bracis-28396 89 29 of of ADP bracis-28396 89 30 pso pso NOUN bracis-28396 89 31 for for ADP bracis-28396 89 32 such such DET bracis-28396 89 33 a a DET bracis-28396 89 34 task task NOUN bracis-28396 89 35 . . PUNCT bracis-28396 90 1 3 3 NUM bracis-28396 90 2 theoretical theoretical ADJ bracis-28396 90 3 background background NOUN bracis-28396 90 4 this this DET bracis-28396 90 5 section section NOUN bracis-28396 90 6 briefly briefly NOUN bracis-28396 90 7 introduces introduce VERB bracis-28396 90 8 the the DET bracis-28396 90 9 main main ADJ bracis-28396 90 10 concepts concept NOUN bracis-28396 90 11 of of ADP bracis-28396 90 12 the the DET bracis-28396 90 13 techniques technique NOUN bracis-28396 90 14 employed employ VERB bracis-28396 90 15 in in ADP bracis-28396 90 16 this this DET bracis-28396 90 17 work work NOUN bracis-28396 90 18 , , PUNCT bracis-28396 90 19 i.e. i.e. X bracis-28396 90 20 , , PUNCT bracis-28396 90 21 the the DET bracis-28396 90 22 multilayer multilayer PROPN bracis-28396 90 23 perceptron perceptron PROPN bracis-28396 90 24 neural neural PROPN bracis-28396 90 25 network network NOUN bracis-28396 90 26 and and CCONJ bracis-28396 90 27 the the DET bracis-28396 90 28 particle particle NOUN bracis-28396 90 29 swarm swarm NOUN bracis-28396 90 30 optimization optimization NOUN bracis-28396 90 31 algorithm algorithm NOUN bracis-28396 90 32 , , PUNCT bracis-28396 90 33 as as ADV bracis-28396 90 34 well well ADV bracis-28396 90 35 as as ADP bracis-28396 90 36 the the DET bracis-28396 90 37 process process NOUN bracis-28396 90 38 of of ADP bracis-28396 90 39 wood wood NOUN bracis-28396 90 40 image image NOUN bracis-28396 90 41 feature feature NOUN bracis-28396 90 42 extraction extraction NOUN bracis-28396 90 43 . . PUNCT bracis-28396 91 1 3.1 3.1 NUM bracis-28396 91 2 multilayer multilayer PROPN bracis-28396 91 3 perceptron perceptron PROPN bracis-28396 91 4 ann ann PROPN bracis-28396 91 5 mlps mlp NOUN bracis-28396 91 6 are be AUX bracis-28396 91 7 composed compose VERB bracis-28396 91 8 of of ADP bracis-28396 91 9 an an DET bracis-28396 91 10 input input NOUN bracis-28396 91 11 and and CCONJ bracis-28396 91 12 an an DET bracis-28396 91 13 output output NOUN bracis-28396 91 14 layer layer NOUN bracis-28396 91 15 , , PUNCT bracis-28396 91 16 as as ADV bracis-28396 91 17 well well ADV bracis-28396 91 18 as as ADP bracis-28396 91 19 one one NUM bracis-28396 91 20 or or CCONJ bracis-28396 91 21 more more ADV bracis-28396 91 22 hidden hidden ADJ bracis-28396 91 23 layers layer NOUN bracis-28396 91 24 of of ADP bracis-28396 91 25 neurons neuron NOUN bracis-28396 91 26 , , PUNCT bracis-28396 91 27 which which PRON bracis-28396 91 28 can can AUX bracis-28396 91 29 be be AUX bracis-28396 91 30 fully fully ADV bracis-28396 91 31 or or CCONJ bracis-28396 91 32 partially partially ADV bracis-28396 91 33 connected connect VERB bracis-28396 91 34 . . PUNCT bracis-28396 92 1 a a DET bracis-28396 92 2 neural neural ADJ bracis-28396 92 3 network network NOUN bracis-28396 92 4 is be AUX bracis-28396 92 5 called call VERB bracis-28396 92 6 fully fully ADV bracis-28396 92 7 connected connect VERB bracis-28396 92 8 when when SCONJ bracis-28396 92 9 each each DET bracis-28396 92 10 neuron neuron NOUN bracis-28396 92 11 from from ADP bracis-28396 92 12 a a DET bracis-28396 92 13 given give VERB bracis-28396 92 14 layer layer NOUN bracis-28396 92 15 is be AUX bracis-28396 92 16 connected connect VERB bracis-28396 92 17 to to ADP bracis-28396 92 18 all all DET bracis-28396 92 19 neurons neuron NOUN bracis-28396 92 20 of of ADP bracis-28396 92 21 the the DET bracis-28396 92 22 next next ADJ bracis-28396 92 23 one one NUM bracis-28396 92 24 . . PUNCT bracis-28396 93 1 similarly similarly ADV bracis-28396 93 2 , , PUNCT bracis-28396 93 3 it it PRON bracis-28396 93 4 is be AUX bracis-28396 93 5 said say VERB bracis-28396 93 6 to to PART bracis-28396 93 7 be be AUX bracis-28396 93 8 partially partially ADV bracis-28396 93 9 connected connect VERB bracis-28396 93 10 when when SCONJ bracis-28396 93 11 some some DET bracis-28396 93 12 neurons neuron NOUN bracis-28396 93 13 of of ADP bracis-28396 93 14 adjacent adjacent ADJ bracis-28396 93 15 layers layer NOUN bracis-28396 93 16 are be AUX bracis-28396 93 17 not not PART bracis-28396 93 18 connected connect VERB bracis-28396 93 19 . . PUNCT bracis-28396 94 1 these these DET bracis-28396 94 2 connections connection NOUN bracis-28396 94 3 are be AUX bracis-28396 94 4 represented represent VERB bracis-28396 94 5 by by ADP bracis-28396 94 6 a a DET bracis-28396 94 7 weight weight NOUN bracis-28396 94 8 matrix matrix NOUN bracis-28396 94 9 , , PUNCT bracis-28396 94 10 which which PRON bracis-28396 94 11 is be AUX bracis-28396 94 12 usually usually ADV bracis-28396 94 13 adjusted adjust VERB bracis-28396 94 14 through through ADP bracis-28396 94 15 gradient gradient NOUN bracis-28396 94 16 - - PUNCT bracis-28396 94 17 based base VERB bracis-28396 94 18 learning learning NOUN bracis-28396 94 19 algorithms algorithm NOUN bracis-28396 94 20 . . PUNCT bracis-28396 95 1 with with ADP bracis-28396 95 2 a a DET bracis-28396 95 3 single single ADJ bracis-28396 95 4 hidden hide VERB bracis-28396 95 5 layer layer NOUN bracis-28396 95 6 , , PUNCT bracis-28396 95 7 an an DET bracis-28396 95 8 mlp mlp NOUN bracis-28396 95 9 is be AUX bracis-28396 95 10 capable capable ADJ bracis-28396 95 11 of of ADP bracis-28396 95 12 representing represent VERB bracis-28396 95 13 a a DET bracis-28396 95 14 large large ADJ bracis-28396 95 15 number number NOUN bracis-28396 95 16 of of ADP bracis-28396 95 17 functions function NOUN bracis-28396 95 18 , , PUNCT bracis-28396 95 19 thus thus ADV bracis-28396 95 20 sufficient sufficient ADJ bracis-28396 95 21 for for ADP bracis-28396 95 22 the the DET bracis-28396 95 23 purpose purpose NOUN bracis-28396 95 24 of of ADP bracis-28396 95 25 this this DET bracis-28396 95 26 study study NOUN bracis-28396 95 27 . . PUNCT bracis-28396 96 1 figure figure NOUN bracis-28396 96 2     SPACE bracis-28396 96 3 2 2 NUM bracis-28396 96 4 depicts depict VERB bracis-28396 96 5 the the DET bracis-28396 96 6 model model NOUN bracis-28396 96 7 architecture architecture NOUN bracis-28396 96 8 . . PUNCT bracis-28396 97 1 fig fig NOUN bracis-28396 97 2 . . PUNCT bracis-28396 98 1 2 2 X bracis-28396 98 2 . . X bracis-28396 98 3 multilayer multilayer PROPN bracis-28396 98 4 perceptron perceptron PROPN bracis-28396 98 5 representation representation PROPN bracis-28396 98 6 . . PUNCT bracis-28396 99 1 full full ADJ bracis-28396 99 2 size size NOUN bracis-28396 99 3 image image NOUN bracis-28396 99 4 the the DET bracis-28396 99 5 conventional conventional ADJ bracis-28396 99 6 algorithm algorithm NOUN bracis-28396 99 7 to to PART bracis-28396 99 8 train train VERB bracis-28396 99 9 an an DET bracis-28396 99 10 mlp mlp NOUN bracis-28396 99 11 network network NOUN bracis-28396 99 12 is be AUX bracis-28396 99 13 the the DET bracis-28396 99 14 backpropagation backpropagation NOUN bracis-28396 99 15 , , PUNCT bracis-28396 99 16 which which PRON bracis-28396 99 17 is be AUX bracis-28396 99 18 composed compose VERB bracis-28396 99 19 of of ADP bracis-28396 99 20 the the DET bracis-28396 99 21 forward forward ADJ bracis-28396 99 22 pass pass NOUN bracis-28396 99 23 , , PUNCT bracis-28396 99 24 which which PRON bracis-28396 99 25 exposes expose VERB bracis-28396 99 26 the the DET bracis-28396 99 27 input input NOUN bracis-28396 99 28 data datum NOUN bracis-28396 99 29 to to ADP bracis-28396 99 30 a a DET bracis-28396 99 31 series series NOUN bracis-28396 99 32 of of ADP bracis-28396 99 33 linear linear PROPN bracis-28396 99 34 operations operation NOUN bracis-28396 99 35 followed follow VERB bracis-28396 99 36 by by ADP bracis-28396 99 37 non non ADJ bracis-28396 99 38 - - ADJ bracis-28396 99 39 linear linear ADJ bracis-28396 99 40 activations activation NOUN bracis-28396 99 41 , , PUNCT bracis-28396 99 42 and and CCONJ bracis-28396 99 43 the the DET bracis-28396 99 44 backward backward ADJ bracis-28396 99 45 phases phase NOUN bracis-28396 99 46 , , PUNCT bracis-28396 99 47 which which PRON bracis-28396 99 48 is be AUX bracis-28396 99 49 responsible responsible ADJ bracis-28396 99 50 for for ADP bracis-28396 99 51 propagating propagate VERB bracis-28396 99 52 the the DET bracis-28396 99 53 output output NOUN bracis-28396 99 54 error error NOUN bracis-28396 99 55 and and CCONJ bracis-28396 99 56 update update VERB bracis-28396 99 57 the the DET bracis-28396 99 58 network network NOUN bracis-28396 99 59 weights weight NOUN bracis-28396 99 60 . . PUNCT bracis-28396 100 1 3.2 3.2 NUM bracis-28396 100 2 particle particle NOUN bracis-28396 100 3 swarm swarm NOUN bracis-28396 100 4 optimization optimization NOUN bracis-28396 100 5 particle particle NOUN bracis-28396 100 6 swarm swarm NOUN bracis-28396 100 7 optimization optimization NOUN bracis-28396 100 8     SPACE bracis-28396 101 1 [ [ X bracis-28396 101 2 11 11 NUM bracis-28396 101 3 ] ] PUNCT bracis-28396 101 4 is be AUX bracis-28396 101 5 a a DET bracis-28396 101 6 global global ADJ bracis-28396 101 7 optimization optimization NOUN bracis-28396 101 8 technique technique NOUN bracis-28396 101 9 based base VERB bracis-28396 101 10 on on ADP bracis-28396 101 11 the the DET bracis-28396 101 12 social social ADJ bracis-28396 101 13 behavior behavior NOUN bracis-28396 101 14 of of ADP bracis-28396 101 15 birds bird NOUN bracis-28396 101 16 , , PUNCT bracis-28396 101 17 fishes fish NOUN bracis-28396 101 18 , , PUNCT bracis-28396 101 19 and and CCONJ bracis-28396 101 20 insects insect NOUN bracis-28396 101 21 , , PUNCT bracis-28396 101 22 among among ADP bracis-28396 101 23 others other NOUN bracis-28396 101 24 . . PUNCT bracis-28396 102 1 the the DET bracis-28396 102 2 method method NOUN bracis-28396 102 3 comprises comprise VERB bracis-28396 102 4 a a DET bracis-28396 102 5 swarm swarm NOUN bracis-28396 102 6 composed compose VERB bracis-28396 102 7 of of ADP bracis-28396 102 8 a a DET bracis-28396 102 9 set set NOUN bracis-28396 102 10 of of ADP bracis-28396 102 11 individuals individual NOUN bracis-28396 102 12 capable capable ADJ bracis-28396 102 13 of of ADP bracis-28396 102 14 sharing share VERB bracis-28396 102 15 information information NOUN bracis-28396 102 16 among among ADP bracis-28396 102 17 themselves themselves PRON bracis-28396 102 18 concerning concern VERB bracis-28396 102 19 their their PRON bracis-28396 102 20 positions position NOUN bracis-28396 102 21 in in ADP bracis-28396 102 22 the the DET bracis-28396 102 23 search search NOUN bracis-28396 102 24 space space NOUN bracis-28396 102 25 , , PUNCT bracis-28396 102 26 as as ADV bracis-28396 102 27 well well ADV bracis-28396 102 28 as as ADP bracis-28396 102 29 the the DET bracis-28396 102 30 relative relative ADJ bracis-28396 102 31 quality quality NOUN bracis-28396 102 32 of of ADP bracis-28396 102 33 this this DET bracis-28396 102 34 position position NOUN bracis-28396 102 35 , , PUNCT bracis-28396 102 36 denoted denote VERB bracis-28396 102 37 by by ADP bracis-28396 102 38 a a DET bracis-28396 102 39 fitness fitness NOUN bracis-28396 102 40 function function NOUN bracis-28396 102 41 . . PUNCT bracis-28396 103 1 in in ADP bracis-28396 103 2 short short ADJ bracis-28396 103 3 , , PUNCT bracis-28396 103 4 each each DET bracis-28396 103 5 pso pso NOUN bracis-28396 103 6 particle particle NOUN bracis-28396 103 7 is be AUX bracis-28396 103 8 represented represent VERB bracis-28396 103 9 by by ADP bracis-28396 103 10 its its PRON bracis-28396 103 11 current current ADJ bracis-28396 103 12 position position NOUN bracis-28396 103 13 , , PUNCT bracis-28396 103 14 velocity velocity NOUN bracis-28396 103 15 , , PUNCT bracis-28396 103 16 and and CCONJ bracis-28396 103 17 the the DET bracis-28396 103 18 best good ADJ bracis-28396 103 19 position position NOUN bracis-28396 103 20 found find VERB bracis-28396 103 21 during during ADP bracis-28396 103 22 the the DET bracis-28396 103 23 training training NOUN bracis-28396 103 24 process process NOUN bracis-28396 103 25 . . PUNCT bracis-28396 104 1 the the DET bracis-28396 104 2 position position NOUN bracis-28396 104 3 of of ADP bracis-28396 104 4 a a DET bracis-28396 104 5 particle particle NOUN bracis-28396 104 6 i i PRON bracis-28396 104 7 is be AUX bracis-28396 104 8 represented represent VERB bracis-28396 104 9 by by ADP bracis-28396 104 10 a a DET bracis-28396 104 11 point point NOUN bracis-28396 104 12 in in ADP bracis-28396 104 13 a a DET bracis-28396 104 14 d d ADJ bracis-28396 104 15 - - ADJ bracis-28396 104 16 dimensional dimensional ADJ bracis-28396 104 17 space space NOUN bracis-28396 104 18 , , PUNCT bracis-28396 104 19 given give VERB bracis-28396 104 20 by by ADP bracis-28396 104 21 \(\varpsi \(\varpsi NOUN bracis-28396 104 22 _ _ NOUN bracis-28396 104 23 { { PUNCT bracis-28396 104 24 i i NOUN bracis-28396 104 25 } } PUNCT bracis-28396 104 26 = = NOUN bracis-28396 104 27 \{\psi \{\psi NOUN bracis-28396 104 28 _ _ PRON bracis-28396 104 29 { { PUNCT bracis-28396 104 30 i1},\psi i1},\psi ADJ bracis-28396 104 31 _ _ PRON bracis-28396 104 32 { { PUNCT bracis-28396 104 33 i2 i2 PROPN bracis-28396 104 34 } } PUNCT bracis-28396 104 35 , , PUNCT bracis-28396 104 36 \ldots \ldots ADP bracis-28396 104 37 , , PUNCT bracis-28396 104 38 \psi \psi PROPN bracis-28396 104 39 _ _ PUNCT bracis-28396 104 40 { { PUNCT bracis-28396 104 41 id}\}\ id}\}\ NOUN bracis-28396 104 42 ) ) PUNCT bracis-28396 104 43 . . PUNCT bracis-28396 105 1 further far ADV bracis-28396 105 2 , , PUNCT bracis-28396 105 3 the the DET bracis-28396 105 4 particle particle NOUN bracis-28396 105 5 velocity velocity NOUN bracis-28396 105 6 is be AUX bracis-28396 105 7 defined define VERB bracis-28396 105 8 by by ADP bracis-28396 105 9 \(\boldsymbol{v}_i=\{v_{i1},v_{i2},\ldots \(\boldsymbol{v}_i=\{v_{i1},v_{i2},\ldot NOUN bracis-28396 105 10 , , PUNCT bracis-28396 105 11 v_{id}\}\ v_{id}\}\ NOUN bracis-28396 105 12 ) ) PUNCT bracis-28396 105 13 , , PUNCT bracis-28396 105 14 and and CCONJ bracis-28396 105 15 finally finally ADV bracis-28396 105 16 the the DET bracis-28396 105 17 best good ADJ bracis-28396 105 18 position position NOUN bracis-28396 105 19 found find VERB bracis-28396 105 20 by by ADP bracis-28396 105 21 this this DET bracis-28396 105 22 particle particle NOUN bracis-28396 105 23 is be AUX bracis-28396 105 24 represented represent VERB bracis-28396 105 25 by by ADP bracis-28396 105 26 \(\boldsymbol{p}_i=\{p_{i1},p_{i2},\ldots \(\boldsymbol{p}_i=\{p_{i1},p_{i2},\ldot NOUN bracis-28396 105 27 , , PUNCT bracis-28396 105 28 p_{id}\}\ p_{id}\}\ NOUN bracis-28396 105 29 ) ) PUNCT bracis-28396 105 30 . . PUNCT bracis-28396 106 1 besides besides SCONJ bracis-28396 106 2 , , PUNCT bracis-28396 106 3 the the DET bracis-28396 106 4 best good ADJ bracis-28396 106 5 position position NOUN bracis-28396 106 6 found find VERB bracis-28396 106 7 among among ADP bracis-28396 106 8 all all DET bracis-28396 106 9 particles particle NOUN bracis-28396 106 10 is be AUX bracis-28396 106 11 represented represent VERB bracis-28396 106 12 by by ADP bracis-28396 106 13 \(\boldsymbol{p}_g\ \(\boldsymbol{p}_g\ PROPN bracis-28396 106 14 ) ) PUNCT bracis-28396 106 15 . . PUNCT bracis-28396 107 1 a a DET bracis-28396 107 2 particle particle NOUN bracis-28396 107 3 will will AUX bracis-28396 107 4 move move VERB bracis-28396 107 5 in in ADP bracis-28396 107 6 a a DET bracis-28396 107 7 particular particular ADJ bracis-28396 107 8 direction direction NOUN bracis-28396 107 9 depending depend VERB bracis-28396 107 10 on on ADP bracis-28396 107 11 its its PRON bracis-28396 107 12 current current ADJ bracis-28396 107 13 position position NOUN bracis-28396 107 14 , , PUNCT bracis-28396 107 15 velocity velocity NOUN bracis-28396 107 16 , , PUNCT bracis-28396 107 17 and and CCONJ bracis-28396 107 18 best good ADJ bracis-28396 107 19 position position NOUN bracis-28396 107 20 . . PUNCT bracis-28396 108 1 additionally additionally ADV bracis-28396 108 2 , , PUNCT bracis-28396 108 3 it it PRON bracis-28396 108 4 also also ADV bracis-28396 108 5 depends depend VERB bracis-28396 108 6 on on ADP bracis-28396 108 7 the the DET bracis-28396 108 8 best good ADJ bracis-28396 108 9 position position NOUN bracis-28396 108 10 found find VERB bracis-28396 108 11 by by ADP bracis-28396 108 12 the the DET bracis-28396 108 13 other other ADJ bracis-28396 108 14 particles particle NOUN bracis-28396 108 15 in in ADP bracis-28396 108 16 the the DET bracis-28396 108 17 swarm swarm NOUN bracis-28396 108 18 . . PUNCT bracis-28396 109 1 therefore therefore ADV bracis-28396 109 2 , , PUNCT bracis-28396 109 3 the the DET bracis-28396 109 4 position position NOUN bracis-28396 109 5 of of ADP bracis-28396 109 6 a a DET bracis-28396 109 7 particle particle NOUN bracis-28396 109 8 \(\psi \(\psi ADP bracis-28396 109 9 _ _ PUNCT bracis-28396 109 10 { { PUNCT bracis-28396 109 11 ij}(t+1 ij}(t+1 PROPN bracis-28396 109 12 ) ) PUNCT bracis-28396 110 1 = = SYM bracis-28396 111 1 \psi \psi NOUN bracis-28396 111 2 _ _ PUNCT bracis-28396 111 3 { { PUNCT bracis-28396 111 4 ij}(t ij}(t NOUN bracis-28396 111 5 ) ) PUNCT bracis-28396 111 6 + + CCONJ bracis-28396 111 7 v_{ij}(t)\ v_{ij}(t)\ NOUN bracis-28396 111 8 ) ) PUNCT bracis-28396 111 9 is be AUX bracis-28396 111 10 computed compute VERB bracis-28396 111 11 for for ADP bracis-28396 111 12 each each DET bracis-28396 111 13 dimension dimension NOUN bracis-28396 111 14 \(j\in \(j\in NUM bracis-28396 111 15 \{1,2,\ldots \{1,2,\ldot NOUN bracis-28396 111 16 , , PUNCT bracis-28396 111 17 d\}\ d\}\ NOUN bracis-28396 111 18 ) ) PUNCT bracis-28396 111 19 at at ADP bracis-28396 111 20 time time NOUN bracis-28396 111 21 step step NOUN bracis-28396 111 22 t. t. PROPN bracis-28396 111 23 further far ADV bracis-28396 111 24 , , PUNCT bracis-28396 111 25 the the DET bracis-28396 111 26 velocity velocity NOUN bracis-28396 111 27 \(v_{ij}(t)\ \(v_{ij}(t)\ PROPN bracis-28396 111 28 ) ) PUNCT bracis-28396 111 29 is be AUX bracis-28396 111 30 updated update VERB bracis-28396 111 31 using use VERB bracis-28396 111 32 the the DET bracis-28396 111 33 following follow VERB bracis-28396 111 34 equation equation NOUN bracis-28396 111 35 : : PUNCT bracis-28396 111 36 $ $ SYM bracis-28396 111 37 $ $ SYM bracis-28396 111 38 \begin{aligned \begin{aligne VERB bracis-28396 111 39 } } PUNCT bracis-28396 111 40 v_{ij}(t+1 v_{ij}(t+1 NUM bracis-28396 111 41 ) ) PUNCT bracis-28396 112 1 = = SYM bracis-28396 112 2 \vargamma \vargamma PROPN bracis-28396 112 3 \cdot \cdot NOUN bracis-28396 112 4 v_{ij}(t v_{ij}(t NOUN bracis-28396 112 5 ) ) PUNCT bracis-28396 113 1 + + CCONJ bracis-28396 113 2 \varphi \varphi PROPN bracis-28396 113 3 _ _ SYM bracis-28396 113 4 1 1 NUM bracis-28396 113 5 \cdot \cdot NOUN bracis-28396 113 6 r_1 r_1 PROPN bracis-28396 113 7 \cdot \cdot PROPN bracis-28396 113 8 ( ( PUNCT bracis-28396 113 9 p_{ij p_{ij PROPN bracis-28396 113 10 } } PUNCT bracis-28396 113 11 x_{ij}(t x_{ij}(t NOUN bracis-28396 113 12 ) ) PUNCT bracis-28396 113 13 ) ) PUNCT bracis-28396 114 1 + + CCONJ bracis-28396 114 2 \varphi \varphi PROPN bracis-28396 114 3 _ _ PROPN bracis-28396 114 4 2 2 NUM bracis-28396 114 5 \cdot \cdot PROPN bracis-28396 114 6 r_2 r_2 PROPN bracis-28396 114 7 \cdot \cdot PROPN bracis-28396 114 8 ( ( PUNCT bracis-28396 114 9 p_{gd p_{gd ADJ bracis-28396 114 10 } } PUNCT bracis-28396 114 11 x_{ij}(t x_{ij}(t NOUN bracis-28396 114 12 ) ) PUNCT bracis-28396 114 13 ) ) PUNCT bracis-28396 114 14 , , PUNCT bracis-28396 114 15 \end{aligned}$$ \end{aligned}$$ X bracis-28396 114 16 ( ( PUNCT bracis-28396 114 17 1 1 NUM bracis-28396 114 18 ) ) PUNCT bracis-28396 114 19 where where SCONJ bracis-28396 114 20 \(\vargamma \(\vargamma PROPN bracis-28396 114 21 \ \ PROPN bracis-28396 114 22 ) ) PUNCT bracis-28396 114 23 denotes denote VERB bracis-28396 114 24 the the DET bracis-28396 114 25 inertial inertial ADJ bracis-28396 114 26 weight weight NOUN bracis-28396 114 27 , , PUNCT bracis-28396 114 28 introduced introduce VERB bracis-28396 114 29 by by ADP bracis-28396 114 30     SPACE bracis-28396 114 31 [ [ X bracis-28396 114 32 23 23 NUM bracis-28396 114 33 ] ] PUNCT bracis-28396 114 34 to to PART bracis-28396 114 35 balance balance VERB bracis-28396 114 36 the the DET bracis-28396 114 37 global global ADJ bracis-28396 114 38 and and CCONJ bracis-28396 114 39 local local ADJ bracis-28396 114 40 search search NOUN bracis-28396 114 41 , , PUNCT bracis-28396 114 42 \(r_1\ \(r_1\ NOUN bracis-28396 114 43 ) ) PUNCT bracis-28396 114 44 and and CCONJ bracis-28396 114 45 \(r_2\ \(r_2\ VERB bracis-28396 114 46 ) ) PUNCT bracis-28396 114 47 are be AUX bracis-28396 114 48 two two NUM bracis-28396 114 49 independent independent ADJ bracis-28396 114 50 values value NOUN bracis-28396 114 51 uniformly uniformly ADV bracis-28396 114 52 distributed distribute VERB bracis-28396 114 53 in in ADP bracis-28396 114 54 the the DET bracis-28396 114 55 range range NOUN bracis-28396 114 56 [ [ X bracis-28396 114 57 0 0 NUM bracis-28396 114 58 , , PUNCT bracis-28396 114 59     SPACE bracis-28396 114 60 1 1 NUM bracis-28396 114 61 ] ] PUNCT bracis-28396 114 62 , , PUNCT bracis-28396 114 63 and and CCONJ bracis-28396 114 64 \(\varphi \(\varphi PUNCT bracis-28396 114 65 _ _ PRON bracis-28396 114 66 1\ 1\ NUM bracis-28396 114 67 ) ) PUNCT bracis-28396 114 68 and and CCONJ bracis-28396 114 69 \(\varphi \(\varphi PUNCT bracis-28396 114 70 _ _ NOUN bracis-28396 114 71 2\ 2\ NUM bracis-28396 114 72 ) ) PUNCT bracis-28396 114 73 are be AUX bracis-28396 114 74 acceleration acceleration NOUN bracis-28396 114 75 constants constant NOUN bracis-28396 114 76 . . PUNCT bracis-28396 115 1 such such DET bracis-28396 115 2 a a DET bracis-28396 115 3 representation representation NOUN bracis-28396 115 4 of of ADP bracis-28396 115 5 a a DET bracis-28396 115 6 particle particle NOUN bracis-28396 115 7 is be AUX bracis-28396 115 8 suitable suitable ADJ bracis-28396 115 9 for for ADP bracis-28396 115 10 hyperparameter hyperparameter NOUN bracis-28396 115 11 fine fine ADV bracis-28396 115 12 - - PUNCT bracis-28396 115 13 tuning tuning NOUN bracis-28396 115 14 , , PUNCT bracis-28396 115 15 which which PRON bracis-28396 115 16 considers consider VERB bracis-28396 115 17 realand realand NOUN bracis-28396 115 18 integer integer NOUN bracis-28396 115 19 - - PUNCT bracis-28396 115 20 valued value VERB bracis-28396 115 21 numbers number NOUN bracis-28396 115 22 . . PUNCT bracis-28396 116 1 however however ADV bracis-28396 116 2 , , PUNCT bracis-28396 116 3 it it PRON bracis-28396 116 4 is be AUX bracis-28396 116 5 not not PART bracis-28396 116 6 adequate adequate ADJ bracis-28396 116 7 for for ADP bracis-28396 116 8 feature feature NOUN bracis-28396 116 9 selection selection NOUN bracis-28396 116 10 tasks task NOUN bracis-28396 116 11 since since SCONJ bracis-28396 116 12 it it PRON bracis-28396 116 13 requires require VERB bracis-28396 116 14 a a DET bracis-28396 116 15 categorical categorical ADJ bracis-28396 116 16 or or CCONJ bracis-28396 116 17 binary binary ADJ bracis-28396 116 18 representation representation NOUN bracis-28396 116 19 . . PUNCT bracis-28396 117 1 therefore therefore ADV bracis-28396 117 2 , , PUNCT bracis-28396 117 3 this this DET bracis-28396 117 4 work work NOUN bracis-28396 117 5 employs employ VERB bracis-28396 117 6 a a DET bracis-28396 117 7 variation variation NOUN bracis-28396 117 8 of of ADP bracis-28396 117 9 the the DET bracis-28396 117 10 method method NOUN bracis-28396 117 11 , , PUNCT bracis-28396 117 12 adapted adapt VERB bracis-28396 117 13 for for ADP bracis-28396 117 14 feature feature NOUN bracis-28396 117 15 selection selection NOUN bracis-28396 117 16 , , PUNCT bracis-28396 117 17 as as SCONJ bracis-28396 117 18 discussed discuss VERB bracis-28396 117 19 in in ADP bracis-28396 117 20 sect sect NOUN bracis-28396 117 21 . . PUNCT bracis-28396 117 22     SPACE bracis-28396 118 1 4 4 NUM bracis-28396 118 2 . . NOUN bracis-28396 118 3 3.3 3.3 NUM bracis-28396 118 4 feature feature NOUN bracis-28396 118 5 extractors extractor NOUN bracis-28396 118 6 haralick haralick PROPN bracis-28396 118 7 et et PROPN bracis-28396 118 8 al al PROPN bracis-28396 118 9 . . PUNCT bracis-28396 118 10     SPACE bracis-28396 119 1 [ [ X bracis-28396 119 2 8 8 NUM bracis-28396 119 3 ] ] PUNCT bracis-28396 119 4 proposed propose VERB bracis-28396 119 5 a a DET bracis-28396 119 6 set set NOUN bracis-28396 119 7 of of ADP bracis-28396 119 8 mathematical mathematical ADJ bracis-28396 119 9 tools tool NOUN bracis-28396 119 10 to to PART bracis-28396 119 11 extract extract VERB bracis-28396 119 12 statistical statistical ADJ bracis-28396 119 13 features feature NOUN bracis-28396 119 14 from from ADP bracis-28396 119 15 images image NOUN bracis-28396 119 16 using use VERB bracis-28396 119 17 a a DET bracis-28396 119 18 gray gray ADJ bracis-28396 119 19 level level NOUN bracis-28396 119 20 co co NOUN bracis-28396 119 21 - - NOUN bracis-28396 119 22 occurrence occurrence ADJ bracis-28396 119 23 matrix matrix NOUN bracis-28396 119 24 . . PUNCT bracis-28396 120 1 the the DET bracis-28396 120 2 glcm glcm PROPN bracis-28396 120 3 is be AUX bracis-28396 120 4 capable capable ADJ bracis-28396 120 5 of of ADP bracis-28396 120 6 describing describe VERB bracis-28396 120 7 the the DET bracis-28396 120 8 frequency frequency NOUN bracis-28396 120 9 of of ADP bracis-28396 120 10 occurrences occurrence NOUN bracis-28396 120 11 in in ADP bracis-28396 120 12 grayscale grayscale NOUN bracis-28396 120 13 transitions transition NOUN bracis-28396 120 14 for for ADP bracis-28396 120 15 an an DET bracis-28396 120 16 image image NOUN bracis-28396 120 17 in in ADP bracis-28396 120 18 a a DET bracis-28396 120 19 pixel pixel NOUN bracis-28396 120 20 - - PUNCT bracis-28396 120 21 by by ADP bracis-28396 120 22 - - PUNCT bracis-28396 120 23 pixel pixel NOUN bracis-28396 120 24 fashion fashion NOUN bracis-28396 120 25 . . PUNCT bracis-28396 121 1 further far ADV bracis-28396 121 2 , , PUNCT bracis-28396 121 3 the the DET bracis-28396 121 4 features feature NOUN bracis-28396 121 5 are be AUX bracis-28396 121 6 extracted extract VERB bracis-28396 121 7 considering consider VERB bracis-28396 121 8 the the DET bracis-28396 121 9 relationship relationship NOUN bracis-28396 121 10 of of ADP bracis-28396 121 11 each each DET bracis-28396 121 12 pixel pixel NOUN bracis-28396 121 13 with with ADP bracis-28396 121 14 its its PRON bracis-28396 121 15 neighbors neighbor NOUN bracis-28396 121 16 over over ADP bracis-28396 121 17 four four NUM bracis-28396 121 18 different different ADJ bracis-28396 121 19 angles angle NOUN bracis-28396 121 20 , , PUNCT bracis-28396 121 21 i.e. i.e. X bracis-28396 121 22 , , PUNCT bracis-28396 121 23 0 0 NUM bracis-28396 121 24 , , PUNCT bracis-28396 121 25 45 45 NUM bracis-28396 121 26 , , PUNCT bracis-28396 121 27 90 90 NUM bracis-28396 121 28 , , PUNCT bracis-28396 121 29 and and CCONJ bracis-28396 121 30 \(135^\circ \(135^\circ PROPN bracis-28396 121 31 \ \ PROPN bracis-28396 121 32 ) ) PUNCT bracis-28396 121 33 , , PUNCT bracis-28396 121 34 where where SCONJ bracis-28396 121 35 0 0 NUM bracis-28396 121 36 and and CCONJ bracis-28396 121 37 \(90^circ\ \(90^circ\ ADJ bracis-28396 121 38 ) ) PUNCT bracis-28396 121 39 are be AUX bracis-28396 121 40 the the DET bracis-28396 121 41 most most ADV bracis-28396 121 42 employed employ VERB bracis-28396 121 43 ones one NOUN bracis-28396 121 44 . . PUNCT bracis-28396 122 1 moreover moreover ADV bracis-28396 122 2 , , PUNCT bracis-28396 122 3 the the DET bracis-28396 122 4 model model NOUN bracis-28396 122 5 is be AUX bracis-28396 122 6 capable capable ADJ bracis-28396 122 7 of of ADP bracis-28396 122 8 extracting extract VERB bracis-28396 122 9 a a DET bracis-28396 122 10 total total NOUN bracis-28396 122 11 of of ADP bracis-28396 122 12 14 14 NUM bracis-28396 122 13 statistical statistical ADJ bracis-28396 122 14 features feature NOUN bracis-28396 122 15 , , PUNCT bracis-28396 122 16 named name VERB bracis-28396 122 17 angular angular ADJ bracis-28396 122 18 second second ADJ bracis-28396 122 19 moment moment NOUN bracis-28396 122 20 , , PUNCT bracis-28396 122 21 contrast contrast NOUN bracis-28396 122 22 , , PUNCT bracis-28396 122 23 correlation correlation NOUN bracis-28396 122 24 , , PUNCT bracis-28396 122 25 variance variance NOUN bracis-28396 122 26 , , PUNCT bracis-28396 122 27 inverse inverse ADJ bracis-28396 122 28 difference difference NOUN bracis-28396 122 29 moment moment NOUN bracis-28396 122 30 , , PUNCT bracis-28396 122 31 sum sum NOUN bracis-28396 122 32 average average NOUN bracis-28396 122 33 , , PUNCT bracis-28396 122 34 sum sum NOUN bracis-28396 122 35 variance variance NOUN bracis-28396 122 36 , , PUNCT bracis-28396 122 37 sum sum PROPN bracis-28396 122 38 entropy entropy PROPN bracis-28396 122 39 , , PUNCT bracis-28396 122 40 entropy entropy PROPN bracis-28396 122 41 , , PUNCT bracis-28396 122 42 difference difference NOUN bracis-28396 122 43 entropy entropy NOUN bracis-28396 122 44 , , PUNCT bracis-28396 122 45 information information NOUN bracis-28396 122 46 measures measure NOUN bracis-28396 122 47 of of ADP bracis-28396 122 48 correlation correlation NOUN bracis-28396 122 49 ( ( PUNCT bracis-28396 122 50 1 1 NUM bracis-28396 122 51 and and CCONJ bracis-28396 122 52 2 2 NUM bracis-28396 122 53 ) ) PUNCT bracis-28396 122 54 and and CCONJ bracis-28396 122 55 the the DET bracis-28396 122 56 maximal maximal ADJ bracis-28396 122 57 correlation correlation NOUN bracis-28396 122 58 coefficient coefficient NOUN bracis-28396 122 59 . . PUNCT bracis-28396 123 1 another another DET bracis-28396 123 2 well well ADV bracis-28396 123 3 - - PUNCT bracis-28396 123 4 known know VERB bracis-28396 123 5 texture texture NOUN bracis-28396 123 6 descriptor descriptor NOUN bracis-28396 123 7 is be AUX bracis-28396 123 8 the the DET bracis-28396 123 9 local local ADJ bracis-28396 123 10 binary binary NOUN bracis-28396 123 11 pattern pattern NOUN bracis-28396 123 12     SPACE bracis-28396 124 1 [ [ X bracis-28396 124 2 13 13 NUM bracis-28396 124 3 ] ] PUNCT bracis-28396 124 4 , , PUNCT bracis-28396 124 5 which which PRON bracis-28396 124 6 converts convert VERB bracis-28396 124 7 the the DET bracis-28396 124 8 image image NOUN bracis-28396 124 9 to to ADP bracis-28396 124 10 a a DET bracis-28396 124 11 gray gray ADJ bracis-28396 124 12 - - PUNCT bracis-28396 124 13 scale scale NOUN bracis-28396 124 14 level level NOUN bracis-28396 124 15 and and CCONJ bracis-28396 124 16 performs perform VERB bracis-28396 124 17 a a DET bracis-28396 124 18 pixel pixel NOUN bracis-28396 124 19 - - PUNCT bracis-28396 124 20 by by ADP bracis-28396 124 21 - - PUNCT bracis-28396 124 22 pixel pixel NOUN bracis-28396 124 23 comparison comparison NOUN bracis-28396 124 24 over over ADP bracis-28396 124 25 the the DET bracis-28396 124 26 entire entire ADJ bracis-28396 124 27 image image NOUN bracis-28396 124 28 considering consider VERB bracis-28396 124 29 a a DET bracis-28396 124 30 selected select VERB bracis-28396 124 31 number number NOUN bracis-28396 124 32 of of ADP bracis-28396 124 33 neighbors neighbor NOUN bracis-28396 124 34 . . PUNCT bracis-28396 125 1 in in ADP bracis-28396 125 2 this this DET bracis-28396 125 3 comparison comparison NOUN bracis-28396 125 4 , , PUNCT bracis-28396 125 5 the the DET bracis-28396 125 6 central central ADJ bracis-28396 125 7 pixel pixel NOUN bracis-28396 125 8 of of ADP bracis-28396 125 9 the the DET bracis-28396 125 10 square square NOUN bracis-28396 125 11 formed form VERB bracis-28396 125 12 by by ADP bracis-28396 125 13 its its PRON bracis-28396 125 14 neighbor neighbor NOUN bracis-28396 125 15 assumes assume VERB bracis-28396 125 16 the the DET bracis-28396 125 17 value value NOUN bracis-28396 125 18 1 1 NUM bracis-28396 125 19 if if SCONJ bracis-28396 125 20 it it PRON bracis-28396 125 21 is be AUX bracis-28396 125 22 greater great ADJ bracis-28396 125 23 or or CCONJ bracis-28396 125 24 equal equal ADJ bracis-28396 125 25 to to ADP bracis-28396 125 26 its its PRON bracis-28396 125 27 neighbors neighbor NOUN bracis-28396 125 28 , , PUNCT bracis-28396 125 29 or or CCONJ bracis-28396 125 30 0 0 NUM bracis-28396 125 31 otherwise otherwise ADV bracis-28396 125 32 . . PUNCT bracis-28396 126 1 such such ADJ bracis-28396 126 2 value value NOUN bracis-28396 126 3 is be AUX bracis-28396 126 4 stored store VERB bracis-28396 126 5 in in ADP bracis-28396 126 6 an an DET bracis-28396 126 7 array array NOUN bracis-28396 126 8 , , PUNCT bracis-28396 126 9 which which PRON bracis-28396 126 10 is be AUX bracis-28396 126 11 further far ADV bracis-28396 126 12 employed employ VERB bracis-28396 126 13 for for ADP bracis-28396 126 14 converting convert VERB bracis-28396 126 15 the the DET bracis-28396 126 16 binary binary ADJ bracis-28396 126 17 intensity intensity PROPN bracis-28396 126 18 code code NOUN bracis-28396 126 19 to to ADP bracis-28396 126 20 a a DET bracis-28396 126 21 decimal decimal ADJ bracis-28396 126 22 value value NOUN bracis-28396 126 23 for for ADP bracis-28396 126 24 the the DET bracis-28396 126 25 pixel pixel NOUN bracis-28396 126 26 at at ADP bracis-28396 126 27 hand hand NOUN bracis-28396 126 28 . . PUNCT bracis-28396 127 1 afterward afterward ADV bracis-28396 127 2 , , PUNCT bracis-28396 127 3 the the DET bracis-28396 127 4 same same ADJ bracis-28396 127 5 computation computation NOUN bracis-28396 127 6 is be AUX bracis-28396 127 7 performed perform VERB bracis-28396 127 8 for for ADP bracis-28396 127 9 all all DET bracis-28396 127 10 pixels pixel NOUN bracis-28396 127 11 compounding compound VERB bracis-28396 127 12 the the DET bracis-28396 127 13 image image NOUN bracis-28396 127 14 . . PUNCT bracis-28396 128 1 further far ADV bracis-28396 128 2 , , PUNCT bracis-28396 128 3 it it PRON bracis-28396 128 4 is be AUX bracis-28396 128 5 generated generate VERB bracis-28396 128 6 a a DET bracis-28396 128 7 histogram histogram NOUN bracis-28396 128 8 of of ADP bracis-28396 128 9 the the DET bracis-28396 128 10 distribution distribution NOUN bracis-28396 128 11 of of ADP bracis-28396 128 12 the the DET bracis-28396 128 13 values value NOUN bracis-28396 128 14 , , PUNCT bracis-28396 128 15 which which PRON bracis-28396 128 16 will will AUX bracis-28396 128 17 compose compose VERB bracis-28396 128 18 the the DET bracis-28396 128 19 final final ADJ bracis-28396 128 20 vector vector NOUN bracis-28396 128 21 describing describe VERB bracis-28396 128 22 the the DET bracis-28396 128 23 image image NOUN bracis-28396 128 24 . . PUNCT bracis-28396 129 1 notice notice VERB bracis-28396 129 2 the the DET bracis-28396 129 3 method method NOUN bracis-28396 129 4 was be AUX bracis-28396 129 5 initially initially ADV bracis-28396 129 6 proposed propose VERB bracis-28396 129 7 with with ADP bracis-28396 129 8 the the DET bracis-28396 129 9 number number NOUN bracis-28396 129 10 of of ADP bracis-28396 129 11 neighborhood neighborhood NOUN bracis-28396 129 12 fixed fix VERB bracis-28396 129 13 in in ADP bracis-28396 129 14 \(3\times \(3\time NOUN bracis-28396 129 15 3\ 3\ NUM bracis-28396 129 16 ) ) PUNCT bracis-28396 129 17 . . PUNCT bracis-28396 130 1 later later ADV bracis-28396 130 2 , , PUNCT bracis-28396 130 3 some some DET bracis-28396 130 4 changes change NOUN bracis-28396 130 5 allowed allow VERB bracis-28396 130 6 lbp lbp NOUN bracis-28396 130 7 to to PART bracis-28396 130 8 deal deal VERB bracis-28396 130 9 with with ADP bracis-28396 130 10 a a DET bracis-28396 130 11 larger large ADJ bracis-28396 130 12 number number NOUN bracis-28396 130 13 of of ADP bracis-28396 130 14 neighborhoods neighborhood NOUN bracis-28396 130 15 , , PUNCT bracis-28396 130 16 resulting result VERB bracis-28396 130 17 in in ADP bracis-28396 130 18 a a DET bracis-28396 130 19 non non ADJ bracis-28396 130 20 - - ADJ bracis-28396 130 21 square square ADJ bracis-28396 130 22 structure structure NOUN bracis-28396 130 23 , , PUNCT bracis-28396 130 24 usually usually ADV bracis-28396 130 25 employing employ VERB bracis-28396 130 26 a a DET bracis-28396 130 27 circular circular ADJ bracis-28396 130 28 pattern pattern NOUN bracis-28396 130 29 since since SCONJ bracis-28396 130 30 it it PRON bracis-28396 130 31 only only ADV bracis-28396 130 32 requires require VERB bracis-28396 130 33 the the DET bracis-28396 130 34 definition definition NOUN bracis-28396 130 35 of of ADP bracis-28396 130 36 the the DET bracis-28396 130 37 radius radius NOUN bracis-28396 130 38 , , PUNCT bracis-28396 130 39 instead instead ADV bracis-28396 130 40 of of ADP bracis-28396 130 41 the the DET bracis-28396 130 42 nxn nxn PROPN bracis-28396 130 43 arrangement arrangement NOUN bracis-28396 130 44 . . PUNCT bracis-28396 131 1 4 4 NUM bracis-28396 131 2 methodology methodology NOUN bracis-28396 131 3 this this DET bracis-28396 131 4 section section NOUN bracis-28396 131 5 presents present VERB bracis-28396 131 6 the the DET bracis-28396 131 7 methodology methodology NOUN bracis-28396 131 8 concerning concern VERB bracis-28396 131 9 the the DET bracis-28396 131 10 material material NOUN bracis-28396 131 11 and and CCONJ bracis-28396 131 12 methods method NOUN bracis-28396 131 13 employed employ VERB bracis-28396 131 14 during during ADP bracis-28396 131 15 the the DET bracis-28396 131 16 experiments experiment NOUN bracis-28396 131 17 . . PUNCT bracis-28396 132 1 it it PRON bracis-28396 132 2 briefly briefly ADV bracis-28396 132 3 describes describe VERB bracis-28396 132 4 the the DET bracis-28396 132 5 datasets dataset NOUN bracis-28396 132 6 , , PUNCT bracis-28396 132 7 the the DET bracis-28396 132 8 process process NOUN bracis-28396 132 9 of of ADP bracis-28396 132 10 feature feature NOUN bracis-28396 132 11 extraction extraction NOUN bracis-28396 132 12 , , PUNCT bracis-28396 132 13 the the DET bracis-28396 132 14 modeling modeling NOUN bracis-28396 132 15 of of ADP bracis-28396 132 16 the the DET bracis-28396 132 17 hyperparameter hyperparameter NOUN bracis-28396 132 18 fine fine ADV bracis-28396 132 19 - - PUNCT bracis-28396 132 20 tuning tune VERB bracis-28396 132 21 and and CCONJ bracis-28396 132 22 feature feature NOUN bracis-28396 132 23 selection selection NOUN bracis-28396 132 24 processes process NOUN bracis-28396 132 25 using use VERB bracis-28396 132 26 pso pso NOUN bracis-28396 132 27 , , PUNCT bracis-28396 132 28 the the DET bracis-28396 132 29 methods method NOUN bracis-28396 132 30 used use VERB bracis-28396 132 31 for for ADP bracis-28396 132 32 evaluation evaluation NOUN bracis-28396 132 33 purpose purpose NOUN bracis-28396 132 34 , , PUNCT bracis-28396 132 35 and and CCONJ bracis-28396 132 36 the the DET bracis-28396 132 37 baselines baseline NOUN bracis-28396 132 38 considered consider VERB bracis-28396 132 39 for for ADP bracis-28396 132 40 comparison comparison NOUN bracis-28396 132 41 . . PUNCT bracis-28396 133 1 4.1 4.1 NUM bracis-28396 133 2 dataset dataset VERB bracis-28396 133 3 this this DET bracis-28396 133 4 work work NOUN bracis-28396 133 5 employs employ VERB bracis-28396 133 6 a a DET bracis-28396 133 7 dataset dataset NOUN bracis-28396 133 8 d d NOUN bracis-28396 133 9 composed compose VERB bracis-28396 133 10 of of ADP bracis-28396 133 11 features feature NOUN bracis-28396 133 12 extracted extract VERB bracis-28396 133 13 from from ADP bracis-28396 133 14 374 374 NUM bracis-28396 133 15 instances instance NOUN bracis-28396 133 16 of of ADP bracis-28396 133 17 wood wood NOUN bracis-28396 133 18 board board NOUN bracis-28396 133 19 images image NOUN bracis-28396 133 20 obtained obtain VERB bracis-28396 133 21 in in ADP bracis-28396 133 22 a a DET bracis-28396 133 23 brazilian brazilian ADJ bracis-28396 133 24 sawmill sawmill NOUN bracis-28396 133 25     SPACE bracis-28396 134 1 [ [ X bracis-28396 134 2 25 25 NUM bracis-28396 134 3 ] ] PUNCT bracis-28396 134 4 . . PUNCT bracis-28396 135 1 as as SCONJ bracis-28396 135 2 stated state VERB bracis-28396 135 3 in in ADP bracis-28396 135 4 the the DET bracis-28396 135 5 problem problem NOUN bracis-28396 135 6 definition definition NOUN bracis-28396 135 7 , , PUNCT bracis-28396 135 8 each each DET bracis-28396 135 9 instance instance NOUN bracis-28396 135 10 \(d=(\boldsymbol{x}_i \(d=(\boldsymbol{x}_i PROPN bracis-28396 135 11 , , PUNCT bracis-28396 135 12 y_i)\ y_i)\ NUM bracis-28396 135 13 ) ) PUNCT bracis-28396 135 14 is be AUX bracis-28396 135 15 composed compose VERB bracis-28396 135 16 of of ADP bracis-28396 135 17 an an DET bracis-28396 135 18 m m ADV bracis-28396 135 19 - - ADJ bracis-28396 135 20 dimensional dimensional ADJ bracis-28396 135 21 feature feature NOUN bracis-28396 135 22 vector vector NOUN bracis-28396 135 23 \(\boldsymbol{x}_i \(\boldsymbol{x}_i PROPN bracis-28396 135 24 \in \in PROPN bracis-28396 135 25 \mathbb \mathbb PROPN bracis-28396 135 26 { { PUNCT bracis-28396 135 27 r}^{m}\ r}^{m}\ NOUN bracis-28396 135 28 ) ) PUNCT bracis-28396 135 29 , , PUNCT bracis-28396 135 30 whose whose DET bracis-28396 135 31 features feature NOUN bracis-28396 135 32 were be AUX bracis-28396 135 33 extracted extract VERB bracis-28396 135 34 using use VERB bracis-28396 135 35 both both DET bracis-28396 135 36 glcm glcm PROPN bracis-28396 135 37 and and CCONJ bracis-28396 135 38 lbp lbp NOUN bracis-28396 135 39 , , PUNCT bracis-28396 135 40 as as SCONJ bracis-28396 135 41 described describe VERB bracis-28396 135 42 in in ADP bracis-28396 135 43 sect sect NOUN bracis-28396 135 44 . . PUNCT bracis-28396 135 45     SPACE bracis-28396 135 46 4.2 4.2 NUM bracis-28396 135 47 , , PUNCT bracis-28396 135 48 and and CCONJ bracis-28396 135 49 a a DET bracis-28396 135 50 target target NOUN bracis-28396 135 51 value value NOUN bracis-28396 135 52 \(y_i\ \(y_i\ NOUN bracis-28396 135 53 ) ) PUNCT bracis-28396 135 54 , , PUNCT bracis-28396 135 55 denoting denote VERB bracis-28396 135 56 the the DET bracis-28396 135 57 wood wood NOUN bracis-28396 135 58 quality quality NOUN bracis-28396 135 59 . . PUNCT bracis-28396 136 1 each each DET bracis-28396 136 2 sample sample NOUN bracis-28396 136 3 ’s ’s PART bracis-28396 136 4 target target NOUN bracis-28396 136 5 value value NOUN bracis-28396 136 6 is be AUX bracis-28396 136 7 established establish VERB bracis-28396 136 8 according accord VERB bracis-28396 136 9 to to ADP bracis-28396 136 10 rules rule NOUN bracis-28396 136 11 defined define VERB bracis-28396 136 12 by by ADP bracis-28396 136 13 the the DET bracis-28396 136 14 sawmill sawmill NOUN bracis-28396 136 15 company company NOUN bracis-28396 136 16 , , PUNCT bracis-28396 136 17 where where SCONJ bracis-28396 136 18 “ " PUNCT bracis-28396 136 19 a a PRON bracis-28396 136 20 ” " PUNCT bracis-28396 136 21 stands stand VERB bracis-28396 136 22 for for ADP bracis-28396 136 23 a a DET bracis-28396 136 24 high high ADJ bracis-28396 136 25 - - PUNCT bracis-28396 136 26 quality quality NOUN bracis-28396 136 27 standard standard NOUN bracis-28396 136 28 and and CCONJ bracis-28396 136 29 comprises comprise VERB bracis-28396 136 30 144 144 NUM bracis-28396 136 31 instances instance NOUN bracis-28396 136 32 , , PUNCT bracis-28396 136 33 “ " PUNCT bracis-28396 136 34 b b X bracis-28396 136 35 ” " PUNCT bracis-28396 136 36 denotes denote VERB bracis-28396 136 37 an an DET bracis-28396 136 38 intermediate intermediate ADJ bracis-28396 136 39 quality quality NOUN bracis-28396 136 40 and and CCONJ bracis-28396 136 41 comprises comprise VERB bracis-28396 136 42 177 177 NUM bracis-28396 136 43 instances instance NOUN bracis-28396 136 44 , , PUNCT bracis-28396 136 45 and and CCONJ bracis-28396 136 46 “ " PUNCT bracis-28396 136 47 c c X bracis-28396 136 48 ” " PUNCT bracis-28396 136 49 represents represent VERB bracis-28396 136 50 lower low ADJ bracis-28396 136 51 quality quality NOUN bracis-28396 136 52 , , PUNCT bracis-28396 136 53 comprising comprise VERB bracis-28396 136 54 53 53 NUM bracis-28396 136 55 samples sample NOUN bracis-28396 136 56 . . PUNCT bracis-28396 137 1 4.2 4.2 NUM bracis-28396 137 2 image image NOUN bracis-28396 137 3 texture texture NOUN bracis-28396 137 4 descriptors descriptor VERB bracis-28396 137 5 the the DET bracis-28396 137 6 feature feature NOUN bracis-28396 137 7 set set NOUN bracis-28396 137 8 was be AUX bracis-28396 137 9 obtained obtain VERB bracis-28396 137 10 by by ADP bracis-28396 137 11 joining join VERB bracis-28396 137 12 the the DET bracis-28396 137 13 features feature NOUN bracis-28396 137 14 extracted extract VERB bracis-28396 137 15 from from ADP bracis-28396 137 16 two two NUM bracis-28396 137 17 texture texture ADJ bracis-28396 137 18 descriptors descriptor NOUN bracis-28396 137 19 , , PUNCT bracis-28396 137 20 namely namely ADV bracis-28396 137 21 , , PUNCT bracis-28396 137 22 statistical statistical ADJ bracis-28396 137 23 measures measure NOUN bracis-28396 137 24 extracted extract VERB bracis-28396 137 25 from from ADP bracis-28396 137 26 glcm glcm NOUN bracis-28396 137 27 and and CCONJ bracis-28396 137 28 the the DET bracis-28396 137 29 lbp lbp NOUN bracis-28396 137 30 . . PUNCT bracis-28396 138 1 while while SCONJ bracis-28396 138 2 the the DET bracis-28396 138 3 statistical statistical ADJ bracis-28396 138 4 measures measure NOUN bracis-28396 138 5 have have VERB bracis-28396 138 6 the the DET bracis-28396 138 7 advantage advantage NOUN bracis-28396 138 8 of of ADP bracis-28396 138 9 enabling enable VERB bracis-28396 138 10 the the DET bracis-28396 138 11 interpretation interpretation NOUN bracis-28396 138 12 and and CCONJ bracis-28396 138 13 comprehension comprehension NOUN bracis-28396 138 14 of of ADP bracis-28396 138 15 the the DET bracis-28396 138 16 image image NOUN bracis-28396 138 17 characteristics characteristic NOUN bracis-28396 138 18 through through ADP bracis-28396 138 19 different different ADJ bracis-28396 138 20 measures measure NOUN bracis-28396 138 21 , , PUNCT bracis-28396 138 22 lbp lbp PROPN bracis-28396 138 23 is be AUX bracis-28396 138 24 robust robust ADJ bracis-28396 138 25 in in ADP bracis-28396 138 26 the the DET bracis-28396 138 27 treatment treatment NOUN bracis-28396 138 28 of of ADP bracis-28396 138 29 gray gray ADJ bracis-28396 138 30 - - PUNCT bracis-28396 138 31 scale scale NOUN bracis-28396 138 32 images image NOUN bracis-28396 138 33 , , PUNCT bracis-28396 138 34 with with ADP bracis-28396 138 35 a a DET bracis-28396 138 36 good good ADJ bracis-28396 138 37 performance performance NOUN bracis-28396 138 38 for for ADP bracis-28396 138 39 scale scale NOUN bracis-28396 138 40 changes change NOUN bracis-28396 138 41 caused cause VERB bracis-28396 138 42 by by ADP bracis-28396 138 43 illumination illumination NOUN bracis-28396 138 44     SPACE bracis-28396 139 1 [ [ X bracis-28396 139 2 13 13 NUM bracis-28396 139 3 ] ] PUNCT bracis-28396 139 4 . . PUNCT bracis-28396 140 1 concerning concern VERB bracis-28396 140 2 the the DET bracis-28396 140 3 glcm glcm NOUN bracis-28396 140 4 , , PUNCT bracis-28396 140 5 this this DET bracis-28396 140 6 paper paper NOUN bracis-28396 140 7 employed employ VERB bracis-28396 140 8 \(0^o\ \(0^o\ NOUN bracis-28396 140 9 ) ) PUNCT bracis-28396 140 10 and and CCONJ bracis-28396 140 11 \(90^o\ \(90^o\ NOUN bracis-28396 140 12 ) ) PUNCT bracis-28396 140 13 to to PART bracis-28396 140 14 extract extract VERB bracis-28396 140 15 six six NUM bracis-28396 140 16 measures measure NOUN bracis-28396 140 17 : : PUNCT bracis-28396 140 18 angular angular ADJ bracis-28396 140 19 second second ADJ bracis-28396 140 20 momentum momentum NOUN bracis-28396 140 21 , , PUNCT bracis-28396 140 22 energy energy NOUN bracis-28396 140 23 , , PUNCT bracis-28396 140 24 contrast contrast NOUN bracis-28396 140 25 , , PUNCT bracis-28396 140 26 correlation correlation NOUN bracis-28396 140 27 , , PUNCT bracis-28396 140 28 dissimilarity dissimilarity NOUN bracis-28396 140 29 , , PUNCT bracis-28396 140 30 and and CCONJ bracis-28396 140 31 homogeneity homogeneity NOUN bracis-28396 140 32 , , PUNCT bracis-28396 140 33 resulting result VERB bracis-28396 140 34 in in ADP bracis-28396 140 35 12 12 NUM bracis-28396 140 36 characteristics characteristic NOUN bracis-28396 140 37 for for ADP bracis-28396 140 38 each each DET bracis-28396 140 39 image image NOUN bracis-28396 140 40 . . PUNCT bracis-28396 141 1 besides besides ADV bracis-28396 141 2 , , PUNCT bracis-28396 141 3 lpb lpb PROPN bracis-28396 141 4 uses use VERB bracis-28396 141 5 24 24 NUM bracis-28396 141 6 neighbors neighbor NOUN bracis-28396 141 7 as as ADV bracis-28396 141 8 well well ADV bracis-28396 141 9 as as ADP bracis-28396 141 10 a a DET bracis-28396 141 11 radius radius NOUN bracis-28396 141 12 of of ADP bracis-28396 141 13 size size NOUN bracis-28396 141 14 3 3 NUM bracis-28396 141 15 , , PUNCT bracis-28396 141 16 resulting result VERB bracis-28396 141 17 in in ADP bracis-28396 141 18 26 26 NUM bracis-28396 141 19 characteristics characteristic NOUN bracis-28396 141 20 for for ADP bracis-28396 141 21 each each DET bracis-28396 141 22 image image NOUN bracis-28396 141 23 . . PUNCT bracis-28396 142 1 thus thus ADV bracis-28396 142 2 , , PUNCT bracis-28396 142 3 joining join VERB bracis-28396 142 4 glcm glcm PROPN bracis-28396 142 5 and and CCONJ bracis-28396 142 6 lbp lbp PROPN bracis-28396 142 7 features feature NOUN bracis-28396 142 8 resulted result VERB bracis-28396 142 9 in in ADP bracis-28396 142 10 38 38 NUM bracis-28396 142 11 predictive predictive ADJ bracis-28396 142 12 attributes attribute NOUN bracis-28396 142 13 . . PUNCT bracis-28396 143 1 4.3 4.3 NUM bracis-28396 143 2 mlp mlp NOUN bracis-28396 143 3 hyperparameter hyperparameter NOUN bracis-28396 143 4 tuning tuning NOUN bracis-28396 143 5 and and CCONJ bracis-28396 143 6     SPACE bracis-28396 143 7 feature feature NOUN bracis-28396 143 8 selection selection NOUN bracis-28396 143 9 using use VERB bracis-28396 143 10 pso pso NOUN bracis-28396 143 11 this this DET bracis-28396 143 12 work work NOUN bracis-28396 143 13 employed employ VERB bracis-28396 143 14 a a DET bracis-28396 143 15 fully fully ADV bracis-28396 143 16 connected connected ADJ bracis-28396 143 17 network network NOUN bracis-28396 143 18 composed compose VERB bracis-28396 143 19 of of ADP bracis-28396 143 20 a a DET bracis-28396 143 21 single single ADJ bracis-28396 143 22 hidden hide VERB bracis-28396 143 23 layer layer NOUN bracis-28396 143 24 for for ADP bracis-28396 143 25 the the DET bracis-28396 143 26 task task NOUN bracis-28396 143 27 of of ADP bracis-28396 143 28 classification classification NOUN bracis-28396 143 29 . . PUNCT bracis-28396 144 1 further far ADV bracis-28396 144 2 , , PUNCT bracis-28396 144 3 it it PRON bracis-28396 144 4 also also ADV bracis-28396 144 5 employed employ VERB bracis-28396 144 6 the the DET bracis-28396 144 7 pso pso NOUN bracis-28396 144 8 algorithm algorithm NOUN bracis-28396 144 9 to to ADP bracis-28396 144 10 fine fine ADV bracis-28396 144 11 - - PUNCT bracis-28396 144 12 tuning tune VERB bracis-28396 144 13 the the DET bracis-28396 144 14 three three NUM bracis-28396 144 15 principal principal ADJ bracis-28396 144 16 hyperparameters hyperparameter NOUN bracis-28396 144 17 of of ADP bracis-28396 144 18 the the DET bracis-28396 144 19 model model NOUN bracis-28396 144 20 , , PUNCT bracis-28396 144 21 namely namely ADV bracis-28396 144 22 the the DET bracis-28396 144 23 number number NOUN bracis-28396 144 24 of of ADP bracis-28396 144 25 neurons neuron NOUN bracis-28396 144 26 in in ADP bracis-28396 144 27 the the DET bracis-28396 144 28 hidden hide VERB bracis-28396 144 29 layer layer NOUN bracis-28396 144 30 \(\gamma \(\gamma X bracis-28396 144 31 \ \ NOUN bracis-28396 144 32 ) ) PUNCT bracis-28396 144 33 , , PUNCT bracis-28396 144 34 the the DET bracis-28396 144 35 learning learn VERB bracis-28396 144 36 rate rate NOUN bracis-28396 144 37 \(\eta \(\eta VERB bracis-28396 144 38 \ \ PROPN bracis-28396 144 39 ) ) PUNCT bracis-28396 144 40 , , PUNCT bracis-28396 144 41 and and CCONJ bracis-28396 144 42 the the DET bracis-28396 144 43 momentum momentum NOUN bracis-28396 144 44 term term NOUN bracis-28396 144 45 \(\mu \(\mu ADP bracis-28396 144 46 \ \ NOUN bracis-28396 144 47 ) ) PUNCT bracis-28396 144 48 . . PUNCT bracis-28396 145 1 as as SCONJ bracis-28396 145 2 mentioned mention VERB bracis-28396 145 3 previously previously ADV bracis-28396 145 4 , , PUNCT bracis-28396 145 5 the the DET bracis-28396 145 6 pso pso NOUN bracis-28396 145 7 algorithm algorithm NOUN bracis-28396 145 8 was be AUX bracis-28396 145 9 employed employ VERB bracis-28396 145 10 to to PART bracis-28396 145 11 perform perform VERB bracis-28396 145 12 a a DET bracis-28396 145 13 combined combine VERB bracis-28396 145 14 task task NOUN bracis-28396 145 15 of of ADP bracis-28396 145 16 hyperparameter hyperparameter NOUN bracis-28396 145 17 tuning tuning NOUN bracis-28396 145 18 and and CCONJ bracis-28396 145 19 feature feature NOUN bracis-28396 145 20 selection selection NOUN bracis-28396 145 21 , , PUNCT bracis-28396 145 22 hereafter hereafter ADV bracis-28396 145 23 referred refer VERB bracis-28396 145 24 to to ADP bracis-28396 145 25 as as ADP bracis-28396 145 26 hp hp PROPN bracis-28396 145 27 - - PUNCT bracis-28396 145 28 fs fs NOUN bracis-28396 145 29 - - PUNCT bracis-28396 145 30 pso pso NOUN bracis-28396 145 31 . . PUNCT bracis-28396 146 1 considering consider VERB bracis-28396 146 2 the the DET bracis-28396 146 3 former former ADJ bracis-28396 146 4 task task NOUN bracis-28396 146 5 , , PUNCT bracis-28396 146 6 pso pso NOUN bracis-28396 146 7 decision decision NOUN bracis-28396 146 8 variables variable NOUN bracis-28396 146 9 are be AUX bracis-28396 146 10 modeled model VERB bracis-28396 146 11 admitting admit VERB bracis-28396 146 12 one one NUM bracis-28396 146 13 integer integer NOUN bracis-28396 146 14 values value NOUN bracis-28396 146 15 to to PART bracis-28396 146 16 represent represent VERB bracis-28396 146 17 the the DET bracis-28396 146 18 number number NOUN bracis-28396 146 19 of of ADP bracis-28396 146 20 units unit NOUN bracis-28396 146 21 in in ADP bracis-28396 146 22 the the DET bracis-28396 146 23 hidden hide VERB bracis-28396 146 24 layer layer NOUN bracis-28396 146 25 , , PUNCT bracis-28396 146 26 as as ADV bracis-28396 146 27 well well ADV bracis-28396 146 28 as as ADP bracis-28396 146 29 two two NUM bracis-28396 146 30 real real ADJ bracis-28396 146 31 numbers number NOUN bracis-28396 146 32 to to PART bracis-28396 146 33 represent represent VERB bracis-28396 146 34 the the DET bracis-28396 146 35 learning learning NOUN bracis-28396 146 36 rate rate NOUN bracis-28396 146 37 and and CCONJ bracis-28396 146 38 momentum momentum NOUN bracis-28396 146 39 . . PUNCT bracis-28396 147 1 further far ADV bracis-28396 147 2 , , PUNCT bracis-28396 147 3 since since SCONJ bracis-28396 147 4 the the DET bracis-28396 147 5 task task NOUN bracis-28396 147 6 of of ADP bracis-28396 147 7 feature feature NOUN bracis-28396 147 8 selection selection NOUN bracis-28396 147 9 does do AUX bracis-28396 147 10 not not PART bracis-28396 147 11 assume assume VERB bracis-28396 147 12 continuous continuous ADJ bracis-28396 147 13 representation representation NOUN bracis-28396 147 14 , , PUNCT bracis-28396 147 15 pso pso NOUN bracis-28396 147 16 requires require VERB bracis-28396 147 17 some some DET bracis-28396 147 18 modifications modification NOUN bracis-28396 147 19 to to PART bracis-28396 147 20 work work VERB bracis-28396 147 21 properly properly ADV bracis-28396 147 22 in in ADP bracis-28396 147 23 this this DET bracis-28396 147 24 context context NOUN bracis-28396 147 25 . . PUNCT bracis-28396 148 1 the the DET bracis-28396 148 2 main main ADJ bracis-28396 148 3 change change NOUN bracis-28396 148 4 is be AUX bracis-28396 148 5 made make VERB bracis-28396 148 6 on on ADP bracis-28396 148 7 the the DET bracis-28396 148 8 position position NOUN bracis-28396 148 9 representation representation NOUN bracis-28396 148 10 , , PUNCT bracis-28396 148 11 which which PRON bracis-28396 148 12 must must AUX bracis-28396 148 13 be be AUX bracis-28396 148 14 treated treat VERB bracis-28396 148 15 as as ADP bracis-28396 148 16 a a DET bracis-28396 148 17 result result NOUN bracis-28396 148 18 of of ADP bracis-28396 148 19 probability probability NOUN bracis-28396 148 20 analysis analysis NOUN bracis-28396 148 21 from from ADP bracis-28396 148 22 particles particle NOUN bracis-28396 148 23 ’ ' PUNCT bracis-28396 148 24 velocity velocity NOUN bracis-28396 148 25 to to PART bracis-28396 148 26 decide decide VERB bracis-28396 148 27 what what PRON bracis-28396 148 28 features feature NOUN bracis-28396 148 29 are be AUX bracis-28396 148 30 relevant relevant ADJ bracis-28396 148 31 to to ADP bracis-28396 148 32 the the DET bracis-28396 148 33 context context NOUN bracis-28396 148 34     SPACE bracis-28396 149 1 [ [ X bracis-28396 149 2 9 9 NUM bracis-28396 149 3 ] ] PUNCT bracis-28396 149 4 . . PUNCT bracis-28396 150 1 therefore therefore ADV bracis-28396 150 2 , , PUNCT bracis-28396 150 3 considering consider VERB bracis-28396 150 4 the the DET bracis-28396 150 5 task task NOUN bracis-28396 150 6 of of ADP bracis-28396 150 7 feature feature NOUN bracis-28396 150 8 selection selection NOUN bracis-28396 150 9 , , PUNCT bracis-28396 150 10 each each DET bracis-28396 150 11 particle particle NOUN bracis-28396 150 12 ’s ’s PART bracis-28396 150 13 decision decision NOUN bracis-28396 150 14 variable variable NOUN bracis-28396 150 15 is be AUX bracis-28396 150 16 represented represent VERB bracis-28396 150 17 by by ADP bracis-28396 150 18 a a DET bracis-28396 150 19 binary binary ADJ bracis-28396 150 20 value value NOUN bracis-28396 150 21 , , PUNCT bracis-28396 150 22 where where SCONJ bracis-28396 150 23 1 1 NUM bracis-28396 150 24 means mean VERB bracis-28396 150 25 to to PART bracis-28396 150 26 consider consider VERB bracis-28396 150 27 a a DET bracis-28396 150 28 feature feature NOUN bracis-28396 150 29 whereas whereas SCONJ bracis-28396 150 30 0 0 NUM bracis-28396 150 31 means mean VERB bracis-28396 150 32 to to PART bracis-28396 150 33 discard discard VERB bracis-28396 150 34 a a DET bracis-28396 150 35 feature feature NOUN bracis-28396 150 36 . . PUNCT bracis-28396 151 1 since since SCONJ bracis-28396 151 2 the the DET bracis-28396 151 3 particle particle NOUN bracis-28396 151 4 ’s ’s PART bracis-28396 151 5 decision decision NOUN bracis-28396 151 6 variables variable NOUN bracis-28396 151 7 employed employ VERB bracis-28396 151 8 for for ADP bracis-28396 151 9 feature feature NOUN bracis-28396 151 10 selection selection NOUN bracis-28396 151 11 assume assume VERB bracis-28396 151 12 binary binary ADJ bracis-28396 151 13 values value NOUN bracis-28396 151 14 , , PUNCT bracis-28396 151 15 it it PRON bracis-28396 151 16 is be AUX bracis-28396 151 17 necessary necessary ADJ bracis-28396 151 18 to to PART bracis-28396 151 19 binarize binarize VERB bracis-28396 151 20 each each DET bracis-28396 151 21 position position NOUN bracis-28396 151 22 , , PUNCT bracis-28396 151 23 \(\psi \(\psi ADP bracis-28396 151 24 _ _ NOUN bracis-28396 151 25 { { PUNCT bracis-28396 151 26 ij}\ ij}\ NOUN bracis-28396 151 27 ) ) PUNCT bracis-28396 152 1 such such ADJ bracis-28396 152 2 that that PRON bracis-28396 152 3 \(\psi \(\psi VERB bracis-28396 152 4 _ _ PUNCT bracis-28396 152 5 { { PUNCT bracis-28396 152 6 ij}=1\ ij}=1\ NOUN bracis-28396 152 7 ) ) PUNCT bracis-28396 152 8 if if SCONJ bracis-28396 152 9 \(s(v_{ij \(s(v_{ij NOUN bracis-28396 152 10 } } PUNCT bracis-28396 152 11 ) ) PUNCT bracis-28396 153 1 > > X bracis-28396 153 2 r_{3}\ r_{3}\ X bracis-28396 153 3 ) ) PUNCT bracis-28396 153 4 and and CCONJ bracis-28396 153 5 0 0 NUM bracis-28396 153 6 otherwise otherwise ADV bracis-28396 153 7 . . PUNCT bracis-28396 154 1 notice notice VERB bracis-28396 154 2 \(r_{3}\ \(r_{3}\ NOUN bracis-28396 154 3 ) ) PUNCT bracis-28396 154 4 is be AUX bracis-28396 154 5 the the DET bracis-28396 154 6 threshold threshold NOUN bracis-28396 154 7 , , PUNCT bracis-28396 154 8 a a DET bracis-28396 154 9 real real ADJ bracis-28396 154 10 number number NOUN bracis-28396 154 11 generated generate VERB bracis-28396 154 12 randomly randomly ADV bracis-28396 154 13 in in ADP bracis-28396 154 14 the the DET bracis-28396 154 15 range range NOUN bracis-28396 154 16 [ [ X bracis-28396 154 17 0 0 NUM bracis-28396 154 18 , , PUNCT bracis-28396 154 19     SPACE bracis-28396 154 20 1 1 NUM bracis-28396 154 21 ] ] PUNCT bracis-28396 154 22 , , PUNCT bracis-28396 154 23 and and CCONJ bracis-28396 154 24 \(s(\cdot \(s(\cdot NOUN bracis-28396 154 25 ) ) PUNCT bracis-28396 155 1 \ \ NOUN bracis-28396 155 2 ) ) PUNCT bracis-28396 155 3 is be AUX bracis-28396 155 4 the the DET bracis-28396 155 5 logistic logistic ADJ bracis-28396 155 6 function function NOUN bracis-28396 155 7 . . PUNCT bracis-28396 156 1 concerning concern VERB bracis-28396 156 2 the the DET bracis-28396 156 3 mlp mlp NOUN bracis-28396 156 4 hyperparameters hyperparameter NOUN bracis-28396 156 5 ’ ' PUNCT bracis-28396 156 6 search search NOUN bracis-28396 156 7 configuration configuration NOUN bracis-28396 156 8 , , PUNCT bracis-28396 156 9 the the DET bracis-28396 156 10 hidden hide VERB bracis-28396 156 11 layer layer NOUN bracis-28396 156 12 size size NOUN bracis-28396 156 13 is be AUX bracis-28396 156 14 optimized optimize VERB bracis-28396 156 15 in in ADP bracis-28396 156 16 the the DET bracis-28396 156 17 range range NOUN bracis-28396 156 18 [ [ X bracis-28396 156 19 2 2 NUM bracis-28396 156 20 , , PUNCT bracis-28396 156 21     SPACE bracis-28396 156 22 60 60 NUM bracis-28396 156 23 ] ] PUNCT bracis-28396 156 24 , , PUNCT bracis-28396 156 25 the the DET bracis-28396 156 26 learning learning NOUN bracis-28396 156 27 rate rate NOUN bracis-28396 156 28 and and CCONJ bracis-28396 156 29 momentum momentum NOUN bracis-28396 156 30 assume assume VERB bracis-28396 156 31 values value NOUN bracis-28396 156 32 in in ADP bracis-28396 156 33 the the DET bracis-28396 156 34 range range NOUN bracis-28396 156 35 [ [ X bracis-28396 156 36 0 0 NUM bracis-28396 156 37 , , PUNCT bracis-28396 156 38     SPACE bracis-28396 156 39 1 1 NUM bracis-28396 156 40 ] ] PUNCT bracis-28396 156 41 , , PUNCT bracis-28396 156 42 and and CCONJ bracis-28396 156 43 the the DET bracis-28396 156 44 feature feature NOUN bracis-28396 156 45 subset subset NOUN bracis-28396 156 46 selection selection NOUN bracis-28396 156 47 is be AUX bracis-28396 156 48 defined define VERB bracis-28396 156 49 by by ADP bracis-28396 156 50 a a DET bracis-28396 156 51 binary binary ADJ bracis-28396 156 52 variable variable NOUN bracis-28396 156 53 , , PUNCT bracis-28396 156 54 i.e. i.e. X bracis-28396 156 55 , , PUNCT bracis-28396 156 56 assuming assume VERB bracis-28396 156 57 either either CCONJ bracis-28396 156 58 0 0 NUM bracis-28396 156 59 or or CCONJ bracis-28396 156 60 1 1 NUM bracis-28396 156 61 . . PUNCT bracis-28396 157 1 additionally additionally ADV bracis-28396 157 2 , , PUNCT bracis-28396 157 3 it it PRON bracis-28396 157 4 also also ADV bracis-28396 157 5 presents present VERB bracis-28396 157 6 the the DET bracis-28396 157 7 mlp mlp PROPN bracis-28396 157 8 hyperparameter hyperparameter NOUN bracis-28396 157 9 default default NOUN bracis-28396 157 10 values value NOUN bracis-28396 157 11 used use VERB bracis-28396 157 12 by by ADP bracis-28396 157 13 weka weka PROPN bracis-28396 157 14     SPACE bracis-28396 158 1 [ [ X bracis-28396 158 2 7 7 NUM bracis-28396 158 3 ] ] PUNCT bracis-28396 158 4 . . PUNCT bracis-28396 159 1 the the DET bracis-28396 159 2 number number NOUN bracis-28396 159 3 of of ADP bracis-28396 159 4 neurons neuron NOUN bracis-28396 159 5 in in ADP bracis-28396 159 6 the the DET bracis-28396 159 7 hidden hide VERB bracis-28396 159 8 layer layer NOUN bracis-28396 159 9 is be AUX bracis-28396 159 10 defined define VERB bracis-28396 159 11 as as ADP bracis-28396 159 12 \(\gamma \(\gamma ADP bracis-28396 160 1 = = PROPN bracis-28396 160 2 \frac{(\text \frac{(\text NOUN bracis-28396 160 3 { { PUNCT bracis-28396 160 4 na na NOUN bracis-28396 160 5 } } PUNCT bracis-28396 160 6 + + CCONJ bracis-28396 160 7 \text \text PROPN bracis-28396 160 8 { { PUNCT bracis-28396 160 9 nc})}{2}\ nc})}{2}\ NOUN bracis-28396 160 10 ) ) PUNCT bracis-28396 160 11 , , PUNCT bracis-28396 160 12 where where SCONJ bracis-28396 160 13 na na PROPN bracis-28396 160 14 is be AUX bracis-28396 160 15 the the DET bracis-28396 160 16 number number NOUN bracis-28396 160 17 of of ADP bracis-28396 160 18 attributes attribute NOUN bracis-28396 160 19 and and CCONJ bracis-28396 160 20 nc nc PROPN bracis-28396 160 21 is be AUX bracis-28396 160 22 the the DET bracis-28396 160 23 number number NOUN bracis-28396 160 24 of of ADP bracis-28396 160 25 classes class NOUN bracis-28396 160 26 . . PUNCT bracis-28396 161 1 thus thus ADV bracis-28396 161 2 , , PUNCT bracis-28396 161 3 the the DET bracis-28396 161 4 default default NOUN bracis-28396 161 5 value value NOUN bracis-28396 161 6 for for ADP bracis-28396 161 7 our our PRON bracis-28396 161 8 problem problem NOUN bracis-28396 161 9 , , PUNCT bracis-28396 161 10 which which PRON bracis-28396 161 11 has have VERB bracis-28396 161 12 38 38 NUM bracis-28396 161 13 attributes attribute NOUN bracis-28396 161 14 and and CCONJ bracis-28396 161 15 3 3 NUM bracis-28396 161 16 classes class NOUN bracis-28396 161 17 , , PUNCT bracis-28396 161 18 is be AUX bracis-28396 161 19 20 20 NUM bracis-28396 161 20 neurons neuron NOUN bracis-28396 161 21 . . PUNCT bracis-28396 162 1 further far ADV bracis-28396 162 2 , , PUNCT bracis-28396 162 3 the the DET bracis-28396 162 4 pso pso NOUN bracis-28396 162 5 algorithm algorithm NOUN bracis-28396 162 6 also also ADV bracis-28396 162 7 has have VERB bracis-28396 162 8 its its PRON bracis-28396 162 9 own own ADJ bracis-28396 162 10 hyperparameters hyperparameter NOUN bracis-28396 162 11 , , PUNCT bracis-28396 162 12 defined define VERB bracis-28396 162 13 as as SCONJ bracis-28396 162 14 follows follow VERB bracis-28396 162 15 : : PUNCT bracis-28396 162 16 number number NOUN bracis-28396 162 17 of of ADP bracis-28396 162 18 particles particle NOUN bracis-28396 162 19 \(n=30\ \(n=30\ NOUN bracis-28396 162 20 ) ) PUNCT bracis-28396 162 21 , , PUNCT bracis-28396 162 22 acceleration acceleration NOUN bracis-28396 162 23 constant constant ADJ bracis-28396 162 24 1 1 NUM bracis-28396 162 25 \(\varphi \(\varphi NOUN bracis-28396 162 26 _ _ PRON bracis-28396 162 27 { { PUNCT bracis-28396 162 28 1}=1.494\ 1}=1.494\ NUM bracis-28396 162 29 ) ) PUNCT bracis-28396 162 30 , , PUNCT bracis-28396 162 31 acceleration acceleration NOUN bracis-28396 162 32 constant constant ADJ bracis-28396 162 33 2 2 NUM bracis-28396 162 34 \(\varphi \(\varphi NOUN bracis-28396 162 35 _ _ PRON bracis-28396 162 36 { { PUNCT bracis-28396 162 37 2}=1.494\ 2}=1.494\ NUM bracis-28396 162 38 ) ) PUNCT bracis-28396 162 39 , , PUNCT bracis-28396 162 40 inertia inertia NOUN bracis-28396 162 41 weight weight NOUN bracis-28396 162 42 \(\omega \(\omega NOUN bracis-28396 162 43 = = NOUN bracis-28396 162 44 0.729\ 0.729\ NOUN bracis-28396 162 45 ) ) PUNCT bracis-28396 162 46 , , PUNCT bracis-28396 162 47 and and CCONJ bracis-28396 162 48 maximum maximum ADJ bracis-28396 162 49 velocity velocity NOUN bracis-28396 162 50 \(\upsilon \(\upsilon NOUN bracis-28396 162 51 \ \ PROPN bracis-28396 162 52 ) ) PUNCT bracis-28396 162 53 . . PUNCT bracis-28396 163 1 notice notice NOUN bracis-28396 163 2 tuning tune VERB bracis-28396 163 3 such such ADJ bracis-28396 163 4 hyperparameters hyperparameter NOUN bracis-28396 163 5 would would AUX bracis-28396 163 6 lead lead VERB bracis-28396 163 7 to to ADP bracis-28396 163 8 a a DET bracis-28396 163 9 “ " PUNCT bracis-28396 163 10 never never ADV bracis-28396 163 11 - - PUNCT bracis-28396 163 12 ending end VERB bracis-28396 163 13 ” " PUNCT bracis-28396 163 14 problem problem NOUN bracis-28396 163 15 . . PUNCT bracis-28396 164 1 therefore therefore ADV bracis-28396 164 2 , , PUNCT bracis-28396 164 3 these these DET bracis-28396 164 4 values value NOUN bracis-28396 164 5 were be AUX bracis-28396 164 6 empirically empirically ADV bracis-28396 164 7 selected select VERB bracis-28396 164 8 based base VERB bracis-28396 164 9 on on ADP bracis-28396 164 10 similar similar ADJ bracis-28396 164 11 works work NOUN bracis-28396 164 12     SPACE bracis-28396 165 1 [ [ X bracis-28396 165 2 9 9 NUM bracis-28396 165 3 , , PUNCT bracis-28396 165 4 21 21 NUM bracis-28396 165 5 ] ] PUNCT bracis-28396 165 6 . . PUNCT bracis-28396 166 1 besides besides SCONJ bracis-28396 166 2 , , PUNCT bracis-28396 166 3 the the DET bracis-28396 166 4 maximum maximum ADJ bracis-28396 166 5 velocity velocity NOUN bracis-28396 166 6 \(\upsilon \(\upsilon NOUN bracis-28396 166 7 \ \ NOUN bracis-28396 166 8 ) ) PUNCT bracis-28396 166 9 varies vary VERB bracis-28396 166 10 according accord VERB bracis-28396 166 11 to to ADP bracis-28396 166 12 the the DET bracis-28396 166 13 upper upper ADJ bracis-28396 166 14 limit limit NOUN bracis-28396 166 15 of of ADP bracis-28396 166 16 its its PRON bracis-28396 166 17 respective respective ADJ bracis-28396 166 18 hyperparameter hyperparameter NOUN bracis-28396 166 19 . . PUNCT bracis-28396 167 1 the the DET bracis-28396 167 2 optimization optimization NOUN bracis-28396 167 3 process process NOUN bracis-28396 167 4 is be AUX bracis-28396 167 5 performed perform VERB bracis-28396 167 6 until until ADP bracis-28396 167 7 meeting meet VERB bracis-28396 167 8 the the DET bracis-28396 167 9 stop stop NOUN bracis-28396 167 10 criterion criterion NOUN bracis-28396 167 11 , , PUNCT bracis-28396 167 12 i.e. i.e. X bracis-28396 167 13 , , PUNCT bracis-28396 167 14 the the DET bracis-28396 167 15 maximum maximum ADJ bracis-28396 167 16 number number NOUN bracis-28396 167 17 of of ADP bracis-28396 167 18 iterations iteration NOUN bracis-28396 167 19 , , PUNCT bracis-28396 167 20 which which PRON bracis-28396 167 21 was be AUX bracis-28396 167 22 set set VERB bracis-28396 167 23 in in ADP bracis-28396 167 24 300 300 NUM bracis-28396 167 25 . . PUNCT bracis-28396 168 1 4.4 4.4 NUM bracis-28396 168 2 evaluation evaluation NOUN bracis-28396 168 3 metaheuristic metaheuristic ADJ bracis-28396 168 4 algorithms algorithm NOUN bracis-28396 168 5 guide guide VERB bracis-28396 168 6 their their PRON bracis-28396 168 7 search search NOUN bracis-28396 168 8 for for ADP bracis-28396 168 9 the the DET bracis-28396 168 10 best good ADJ bracis-28396 168 11 solutions solution NOUN bracis-28396 168 12 according accord VERB bracis-28396 168 13 to to ADP bracis-28396 168 14 the the DET bracis-28396 168 15 outcome outcome NOUN bracis-28396 168 16 of of ADP bracis-28396 168 17 a a DET bracis-28396 168 18 fitness fitness NOUN bracis-28396 168 19 function function NOUN bracis-28396 168 20 . . PUNCT bracis-28396 169 1 in in ADP bracis-28396 169 2 this this DET bracis-28396 169 3 work work NOUN bracis-28396 169 4 , , PUNCT bracis-28396 169 5 this this DET bracis-28396 169 6 measure measure NOUN bracis-28396 169 7 is be AUX bracis-28396 169 8 obtained obtain VERB bracis-28396 169 9 from from ADP bracis-28396 169 10 the the DET bracis-28396 169 11 predictive predictive ADJ bracis-28396 169 12 performance performance NOUN bracis-28396 169 13 of of ADP bracis-28396 169 14 the the DET bracis-28396 169 15 anns anns NOUN bracis-28396 169 16 over over ADP bracis-28396 169 17 a a DET bracis-28396 169 18 validation validation NOUN bracis-28396 169 19 subset subset NOUN bracis-28396 169 20 . . PUNCT bracis-28396 170 1 moreover moreover ADV bracis-28396 170 2 , , PUNCT bracis-28396 170 3 since since SCONJ bracis-28396 170 4 the the DET bracis-28396 170 5 dataset dataset NOUN bracis-28396 170 6 used use VERB bracis-28396 170 7 in in ADP bracis-28396 170 8 this this DET bracis-28396 170 9 work work NOUN bracis-28396 170 10 is be AUX bracis-28396 170 11 composed compose VERB bracis-28396 170 12 of of ADP bracis-28396 170 13 an an DET bracis-28396 170 14 imbalanced imbalanced ADJ bracis-28396 170 15 class class NOUN bracis-28396 170 16 distribution distribution NOUN bracis-28396 170 17 , , PUNCT bracis-28396 170 18 i.e. i.e. X bracis-28396 170 19 , , PUNCT bracis-28396 170 20 there there PRON bracis-28396 170 21 is be VERB bracis-28396 170 22 a a DET bracis-28396 170 23 considerably considerably ADV bracis-28396 170 24 smaller small ADJ bracis-28396 170 25 number number NOUN bracis-28396 170 26 of of ADP bracis-28396 170 27 examples example NOUN bracis-28396 170 28 labeled label VERB bracis-28396 170 29 as as ADP bracis-28396 170 30 “ " PUNCT bracis-28396 170 31 c c NOUN bracis-28396 170 32 ” " PUNCT bracis-28396 170 33 class class NOUN bracis-28396 170 34 . . PUNCT bracis-28396 171 1 therefore therefore ADV bracis-28396 171 2 , , PUNCT bracis-28396 171 3 measures measure NOUN bracis-28396 171 4 that that PRON bracis-28396 171 5 acknowledge acknowledge VERB bracis-28396 171 6 this this DET bracis-28396 171 7 imbalance imbalance NOUN bracis-28396 171 8 are be AUX bracis-28396 171 9 more more ADV bracis-28396 171 10 suitable suitable ADJ bracis-28396 171 11 for for ADP bracis-28396 171 12 the the DET bracis-28396 171 13 task task NOUN bracis-28396 171 14 . . PUNCT bracis-28396 172 1 thus thus ADV bracis-28396 172 2 , , PUNCT bracis-28396 172 3 in in ADP bracis-28396 172 4 this this DET bracis-28396 172 5 study study NOUN bracis-28396 172 6 , , PUNCT bracis-28396 172 7 we we PRON bracis-28396 172 8 considered consider VERB bracis-28396 172 9 the the DET bracis-28396 172 10 balanced balanced ADJ bracis-28396 172 11 accuracy accuracy NOUN bracis-28396 172 12 ( ( PUNCT bracis-28396 172 13 bac bac NOUN bracis-28396 172 14 ) ) PUNCT bracis-28396 172 15     SPACE bracis-28396 173 1 [ [ X bracis-28396 173 2 3 3 NUM bracis-28396 173 3 ] ] PUNCT bracis-28396 173 4 . . PUNCT bracis-28396 174 1 the the DET bracis-28396 174 2 ann ann PROPN bracis-28396 174 3 training training NOUN bracis-28396 174 4 and and CCONJ bracis-28396 174 5 evaluation evaluation NOUN bracis-28396 174 6 were be AUX bracis-28396 174 7 performed perform VERB bracis-28396 174 8 using use VERB bracis-28396 174 9 a a DET bracis-28396 174 10 nested nest VERB bracis-28396 174 11 stratified stratify VERB bracis-28396 174 12 k k ADJ bracis-28396 174 13 - - ADJ bracis-28396 174 14 fold fold ADJ bracis-28396 174 15 cross cross NOUN bracis-28396 174 16 - - NOUN bracis-28396 174 17 validation validation ADJ bracis-28396 174 18 ( ( PUNCT bracis-28396 174 19 cv cv NOUN bracis-28396 174 20 ) ) PUNCT bracis-28396 174 21 re re ADJ bracis-28396 174 22 - - ADJ bracis-28396 174 23 sampling sample VERB bracis-28396 174 24 method method NOUN bracis-28396 174 25 . . PUNCT bracis-28396 175 1 such such DET bracis-28396 175 2 an an DET bracis-28396 175 3 approach approach NOUN bracis-28396 175 4 splits split VERB bracis-28396 175 5 the the DET bracis-28396 175 6 dataset dataset NOUN bracis-28396 175 7 into into ADP bracis-28396 175 8 \(k=10\ \(k=10\ PROPN bracis-28396 175 9 ) ) PUNCT bracis-28396 175 10 partitions partition NOUN bracis-28396 175 11 , , PUNCT bracis-28396 175 12 where where SCONJ bracis-28396 175 13 one one NUM bracis-28396 175 14 of of ADP bracis-28396 175 15 them they PRON bracis-28396 175 16 is be AUX bracis-28396 175 17 used use VERB bracis-28396 175 18 to to PART bracis-28396 175 19 test test VERB bracis-28396 175 20 , , PUNCT bracis-28396 175 21 and and CCONJ bracis-28396 175 22 the the DET bracis-28396 175 23 remaining remain VERB bracis-28396 175 24 folds fold NOUN bracis-28396 175 25 are be AUX bracis-28396 175 26 employed employ VERB bracis-28396 175 27 for for ADP bracis-28396 175 28 training training NOUN bracis-28396 175 29 purposes purpose NOUN bracis-28396 175 30 . . PUNCT bracis-28396 176 1 in in ADP bracis-28396 176 2 this this DET bracis-28396 176 3 context context NOUN bracis-28396 176 4 , , PUNCT bracis-28396 176 5 the the DET bracis-28396 176 6 training training NOUN bracis-28396 176 7 folds fold NOUN bracis-28396 176 8 are be AUX bracis-28396 176 9 used use VERB bracis-28396 176 10 to to PART bracis-28396 176 11 train train VERB bracis-28396 176 12 the the DET bracis-28396 176 13 model model NOUN bracis-28396 176 14 during during ADP bracis-28396 176 15 the the DET bracis-28396 176 16 optimization optimization NOUN bracis-28396 176 17 process process NOUN bracis-28396 176 18 , , PUNCT bracis-28396 176 19 i.e. i.e. X bracis-28396 176 20 , , PUNCT bracis-28396 176 21 finding find VERB bracis-28396 176 22 the the DET bracis-28396 176 23 best good ADJ bracis-28396 176 24 mlp mlp NOUN bracis-28396 176 25 hyperparameters hyperparameter NOUN bracis-28396 176 26 and and CCONJ bracis-28396 176 27 the the DET bracis-28396 176 28 best good ADJ bracis-28396 176 29 subset subset NOUN bracis-28396 176 30 of of ADP bracis-28396 176 31 features feature NOUN bracis-28396 176 32 . . PUNCT bracis-28396 177 1 therefore therefore ADV bracis-28396 177 2 , , PUNCT bracis-28396 177 3 pso pso NOUN bracis-28396 177 4 assesses assess VERB bracis-28396 177 5 the the DET bracis-28396 177 6 average average ADJ bracis-28396 177 7 bac bac NOUN bracis-28396 177 8 considering consider VERB bracis-28396 177 9 the the DET bracis-28396 177 10 fitness fitness NOUN bracis-28396 177 11 value value NOUN bracis-28396 177 12 over over ADP bracis-28396 177 13 the the DET bracis-28396 177 14 validation validation NOUN bracis-28396 177 15 set set NOUN bracis-28396 177 16 . . PUNCT bracis-28396 178 1 further far ADV bracis-28396 178 2 , , PUNCT bracis-28396 178 3 the the DET bracis-28396 178 4 best good ADJ bracis-28396 178 5 set set NOUN bracis-28396 178 6 of of ADP bracis-28396 178 7 hyperparameters hyperparameter NOUN bracis-28396 178 8 and and CCONJ bracis-28396 178 9 features feature NOUN bracis-28396 178 10 found find VERB bracis-28396 178 11 in in ADP bracis-28396 178 12 this this DET bracis-28396 178 13 process process NOUN bracis-28396 178 14 is be AUX bracis-28396 178 15 them they PRON bracis-28396 178 16 applied apply VERB bracis-28396 178 17 to to PART bracis-28396 178 18 train train VERB bracis-28396 178 19 the the DET bracis-28396 178 20 model model NOUN bracis-28396 178 21 and and CCONJ bracis-28396 178 22 induct induct PROPN bracis-28396 178 23 the the DET bracis-28396 178 24 prediction prediction NOUN bracis-28396 178 25 of of ADP bracis-28396 178 26 the the DET bracis-28396 178 27 testing testing NOUN bracis-28396 178 28 set set VERB bracis-28396 178 29 samples sample NOUN bracis-28396 178 30 ’ ' PUNCT bracis-28396 178 31 labels label NOUN bracis-28396 178 32 . . PUNCT bracis-28396 179 1 such such DET bracis-28396 179 2 a a DET bracis-28396 179 3 process process NOUN bracis-28396 179 4 guarantees guarantee VERB bracis-28396 179 5 that that SCONJ bracis-28396 179 6 the the DET bracis-28396 179 7 data datum NOUN bracis-28396 179 8 used use VERB bracis-28396 179 9 to to PART bracis-28396 179 10 evaluate evaluate VERB bracis-28396 179 11 the the DET bracis-28396 179 12 model model NOUN bracis-28396 179 13 is be AUX bracis-28396 179 14 never never ADV bracis-28396 179 15 used use VERB bracis-28396 179 16 in in ADP bracis-28396 179 17 the the DET bracis-28396 179 18 model model NOUN bracis-28396 179 19 training training NOUN bracis-28396 179 20 steps step NOUN bracis-28396 179 21 and and CCONJ bracis-28396 179 22 , , PUNCT bracis-28396 179 23 consequently consequently ADV bracis-28396 179 24 , , PUNCT bracis-28396 179 25 in in ADP bracis-28396 179 26 the the DET bracis-28396 179 27 mlp mlp NOUN bracis-28396 179 28 hyperparameter hyperparameter NOUN bracis-28396 179 29 tuning tuning NOUN bracis-28396 179 30 and and CCONJ bracis-28396 179 31 feature feature NOUN bracis-28396 179 32 selection selection NOUN bracis-28396 179 33 processes process NOUN bracis-28396 179 34 . . PUNCT bracis-28396 180 1 finally finally ADV bracis-28396 180 2 , , PUNCT bracis-28396 180 3 due due ADP bracis-28396 180 4 to to ADP bracis-28396 180 5 the the DET bracis-28396 180 6 stochastic stochastic ADJ bracis-28396 180 7 process process NOUN bracis-28396 180 8 of of ADP bracis-28396 180 9 pso pso NOUN bracis-28396 180 10 , , PUNCT bracis-28396 180 11 the the DET bracis-28396 180 12 optimization optimization NOUN bracis-28396 180 13 process process NOUN bracis-28396 180 14 was be AUX bracis-28396 180 15 repeated repeat VERB bracis-28396 180 16 during during ADP bracis-28396 180 17 10 10 NUM bracis-28396 180 18 runs run NOUN bracis-28396 180 19 , , PUNCT bracis-28396 180 20 aiming aim VERB bracis-28396 180 21 to to PART bracis-28396 180 22 perform perform VERB bracis-28396 180 23 a a DET bracis-28396 180 24 statistical statistical ADJ bracis-28396 180 25 analysis analysis NOUN bracis-28396 180 26 through through ADP bracis-28396 180 27 the the DET bracis-28396 180 28 wilcoxon wilcoxon ADJ bracis-28396 180 29 signed sign VERB bracis-28396 180 30 - - PUNCT bracis-28396 180 31 rank rank NOUN bracis-28396 180 32 test test NOUN bracis-28396 180 33     SPACE bracis-28396 181 1 [ [ X bracis-28396 181 2 26 26 NUM bracis-28396 181 3 ] ] PUNCT bracis-28396 181 4 with with ADP bracis-28396 181 5 \(5\%\ \(5\%\ NOUN bracis-28396 181 6 ) ) PUNCT bracis-28396 181 7 of of ADP bracis-28396 181 8 significance significance NOUN bracis-28396 181 9 . . PUNCT bracis-28396 182 1 4.5 4.5 NUM bracis-28396 182 2 baselines baseline NOUN bracis-28396 182 3 in in ADP bracis-28396 182 4 order order NOUN bracis-28396 182 5 to to PART bracis-28396 182 6 evaluate evaluate VERB bracis-28396 182 7 and and CCONJ bracis-28396 182 8 compare compare VERB bracis-28396 182 9 the the DET bracis-28396 182 10 results result NOUN bracis-28396 182 11 obtained obtain VERB bracis-28396 182 12 by by ADP bracis-28396 182 13 pso pso NOUN bracis-28396 182 14 , , PUNCT bracis-28396 182 15 six six NUM bracis-28396 182 16 baselines baseline NOUN bracis-28396 182 17 methods method NOUN bracis-28396 182 18 were be AUX bracis-28396 182 19 compared compare VERB bracis-28396 182 20 in in ADP bracis-28396 182 21 the the DET bracis-28396 182 22 context context NOUN bracis-28396 182 23 of of ADP bracis-28396 182 24 the the DET bracis-28396 182 25 combined combined ADJ bracis-28396 182 26 mlp mlp PROPN bracis-28396 182 27 hyperparameter hyperparameter NOUN bracis-28396 182 28 tuning tuning NOUN bracis-28396 182 29 and and CCONJ bracis-28396 182 30 feature feature NOUN bracis-28396 182 31 selection selection NOUN bracis-28396 182 32 : : PUNCT bracis-28396 182 33 method method NOUN bracis-28396 182 34 1 1 NUM bracis-28396 182 35 ( ( PUNCT bracis-28396 182 36 m1 m1 NOUN bracis-28396 182 37 ) ) PUNCT bracis-28396 182 38 : : PUNCT bracis-28396 182 39 default default NOUN bracis-28396 182 40 hyperparameter hyperparameter NOUN bracis-28396 182 41 values value NOUN bracis-28396 182 42 defined define VERB bracis-28396 182 43 by by ADP bracis-28396 182 44 weka weka NOUN bracis-28396 182 45 and and CCONJ bracis-28396 182 46 the the DET bracis-28396 182 47 whole whole ADJ bracis-28396 182 48 set set NOUN bracis-28396 182 49 of of ADP bracis-28396 182 50 features feature NOUN bracis-28396 182 51 ; ; PUNCT bracis-28396 182 52 method method NOUN bracis-28396 182 53 2 2 NUM bracis-28396 182 54 ( ( PUNCT bracis-28396 182 55 m2 m2 PROPN bracis-28396 182 56 ) ) PUNCT bracis-28396 182 57 : : PUNCT bracis-28396 182 58 mlp mlp PROPN bracis-28396 182 59 hyperparameter hyperparameter NOUN bracis-28396 182 60 tuned tune VERB bracis-28396 182 61 using use VERB bracis-28396 182 62 pso pso NOUN bracis-28396 182 63 and and CCONJ bracis-28396 182 64 the the DET bracis-28396 182 65 whole whole ADJ bracis-28396 182 66 set set NOUN bracis-28396 182 67 of of ADP bracis-28396 182 68 features feature NOUN bracis-28396 182 69 ; ; PUNCT bracis-28396 182 70 method method NOUN bracis-28396 182 71 3 3 NUM bracis-28396 182 72 ( ( PUNCT bracis-28396 182 73 m3 m3 PROPN bracis-28396 182 74 ) ) PUNCT bracis-28396 182 75 : : PUNCT bracis-28396 182 76 default default NOUN bracis-28396 182 77 mlp mlp PROPN bracis-28396 182 78 hyperparameter hyperparameter NOUN bracis-28396 182 79 values value NOUN bracis-28396 182 80 defined define VERB bracis-28396 182 81 by by ADP bracis-28396 182 82 weka weka NOUN bracis-28396 182 83 and and CCONJ bracis-28396 182 84 feature feature NOUN bracis-28396 182 85 subset subset NOUN bracis-28396 182 86 selected select VERB bracis-28396 182 87 by by ADP bracis-28396 182 88 pso pso NOUN bracis-28396 182 89 ; ; PUNCT bracis-28396 182 90 method method NOUN bracis-28396 182 91 4 4 NUM bracis-28396 182 92 ( ( PUNCT bracis-28396 182 93 m4 m4 PROPN bracis-28396 182 94 ) ) PUNCT bracis-28396 182 95 : : PUNCT bracis-28396 182 96 default default NOUN bracis-28396 182 97 mlp mlp PROPN bracis-28396 182 98 hyperparameter hyperparameter NOUN bracis-28396 182 99 values value NOUN bracis-28396 182 100 defined define VERB bracis-28396 182 101 by by ADP bracis-28396 182 102 weka weka NOUN bracis-28396 182 103 and and CCONJ bracis-28396 182 104 dimensionality dimensionality NOUN bracis-28396 182 105 reduction reduction NOUN bracis-28396 182 106 performed perform VERB bracis-28396 182 107 by by ADP bracis-28396 182 108 principal principal ADJ bracis-28396 182 109 components component NOUN bracis-28396 182 110 analysis analysis NOUN bracis-28396 182 111 ( ( PUNCT bracis-28396 182 112 pca pca NOUN bracis-28396 182 113 ) ) PUNCT bracis-28396 182 114 ; ; PUNCT bracis-28396 182 115 method method NOUN bracis-28396 182 116 5 5 NUM bracis-28396 182 117 ( ( PUNCT bracis-28396 182 118 m5 m5 PROPN bracis-28396 182 119 ) ) PUNCT bracis-28396 182 120 : : PUNCT bracis-28396 182 121 mlp mlp NOUN bracis-28396 182 122 hyperparameter hyperparameter NOUN bracis-28396 182 123 values value NOUN bracis-28396 182 124 tuned tune VERB bracis-28396 182 125 by by ADP bracis-28396 182 126 pso pso NOUN bracis-28396 182 127 and and CCONJ bracis-28396 182 128 dimensionality dimensionality NOUN bracis-28396 182 129 reduction reduction NOUN bracis-28396 182 130 performed perform VERB bracis-28396 182 131 by by ADP bracis-28396 182 132 pca pca PROPN bracis-28396 182 133 ; ; PUNCT bracis-28396 182 134 random random ADJ bracis-28396 182 135 search search NOUN bracis-28396 182 136 ( ( PUNCT bracis-28396 182 137 rs rs NOUN bracis-28396 182 138 ) ) PUNCT bracis-28396 182 139 : : PUNCT bracis-28396 182 140 random random ADJ bracis-28396 182 141 selection selection NOUN bracis-28396 182 142 of of ADP bracis-28396 182 143 mlp mlp PROPN bracis-28396 182 144 hyperparameter hyperparameter NOUN bracis-28396 182 145 values value NOUN bracis-28396 182 146 and and CCONJ bracis-28396 182 147 feature feature NOUN bracis-28396 182 148 subset subset NOUN bracis-28396 182 149 . . PUNCT bracis-28396 183 1 this this DET bracis-28396 183 2 approach approach NOUN bracis-28396 183 3 considered consider VERB bracis-28396 183 4 the the DET bracis-28396 183 5 same same ADJ bracis-28396 183 6 number number NOUN bracis-28396 183 7 of of ADP bracis-28396 183 8 solutions solution NOUN bracis-28396 183 9 evaluated evaluate VERB bracis-28396 183 10 by by ADP bracis-28396 183 11 pso pso NOUN bracis-28396 183 12 . . PUNCT bracis-28396 184 1 the the DET bracis-28396 184 2 methodology methodology NOUN bracis-28396 184 3 adopted adopt VERB bracis-28396 184 4 in in ADP bracis-28396 184 5 this this DET bracis-28396 184 6 work work NOUN bracis-28396 184 7 aims aim VERB bracis-28396 184 8 at at ADP bracis-28396 184 9 analyzing analyze VERB bracis-28396 184 10 the the DET bracis-28396 184 11 pso pso NOUN bracis-28396 184 12 performance performance NOUN bracis-28396 184 13 from from ADP bracis-28396 184 14 different different ADJ bracis-28396 184 15 x x ADJ bracis-28396 184 16 perspectives perspective NOUN bracis-28396 184 17 . . PUNCT bracis-28396 185 1 first first ADV bracis-28396 185 2 , , PUNCT bracis-28396 185 3 m1 m1 PROPN bracis-28396 185 4 is be AUX bracis-28396 185 5 the the DET bracis-28396 185 6 baseline baseline NOUN bracis-28396 185 7 for for ADP bracis-28396 185 8 both both DET bracis-28396 185 9 tasks task NOUN bracis-28396 185 10 , , PUNCT bracis-28396 185 11 i.e. i.e. X bracis-28396 185 12 , , PUNCT bracis-28396 185 13 the the DET bracis-28396 185 14 mlp mlp PROPN bracis-28396 185 15 hyperparameter hyperparameter NOUN bracis-28396 185 16 using use VERB bracis-28396 185 17 default default NOUN bracis-28396 185 18 parameters parameter NOUN bracis-28396 185 19 provided provide VERB bracis-28396 185 20 by by ADP bracis-28396 185 21 weka weka NOUN bracis-28396 185 22 considering consider VERB bracis-28396 185 23 the the DET bracis-28396 185 24 whole whole ADJ bracis-28396 185 25 set set NOUN bracis-28396 185 26 of of ADP bracis-28396 185 27 features feature NOUN bracis-28396 185 28 . . PUNCT bracis-28396 186 1 further far ADV bracis-28396 186 2 , , PUNCT bracis-28396 186 3 m2 m2 PROPN bracis-28396 186 4 allows allow VERB bracis-28396 186 5 analyzing analyze VERB bracis-28396 186 6 pso pso NOUN bracis-28396 186 7 influence influence NOUN bracis-28396 186 8 for for ADP bracis-28396 186 9 the the DET bracis-28396 186 10 task task NOUN bracis-28396 186 11 of of ADP bracis-28396 186 12 mlp mlp PROPN bracis-28396 186 13 hyperparameter hyperparameter NOUN bracis-28396 186 14 tuning tuning NOUN bracis-28396 186 15 , , PUNCT bracis-28396 186 16 with with ADP bracis-28396 186 17 no no DET bracis-28396 186 18 feature feature NOUN bracis-28396 186 19 selection selection NOUN bracis-28396 186 20 , , PUNCT bracis-28396 186 21 while while SCONJ bracis-28396 186 22 m3 m3 PROPN bracis-28396 186 23 investigates investigate VERB bracis-28396 186 24 the the DET bracis-28396 186 25 opposite opposite ADJ bracis-28396 186 26 , , PUNCT bracis-28396 186 27 i.e. i.e. X bracis-28396 186 28 , , PUNCT bracis-28396 186 29 pso pso NOUN bracis-28396 186 30 influence influence NOUN bracis-28396 186 31 to to ADP bracis-28396 186 32 the the DET bracis-28396 186 33 task task NOUN bracis-28396 186 34 of of ADP bracis-28396 186 35 feature feature NOUN bracis-28396 186 36 selection selection NOUN bracis-28396 186 37 with with ADP bracis-28396 186 38 using use VERB bracis-28396 186 39 mlp mlp NOUN bracis-28396 186 40 default default NOUN bracis-28396 186 41 hyperparameters hyperparameter NOUN bracis-28396 186 42 . . PUNCT bracis-28396 187 1 moreover moreover ADV bracis-28396 187 2 , , PUNCT bracis-28396 187 3 m4 m4 PROPN bracis-28396 187 4 employs employ VERB bracis-28396 187 5 the the DET bracis-28396 187 6 default default NOUN bracis-28396 187 7 hyperparameter hyperparameter NOUN bracis-28396 187 8 with with ADP bracis-28396 187 9 a a DET bracis-28396 187 10 dimensionally dimensionally ADV bracis-28396 187 11 reduced reduce VERB bracis-28396 187 12 feature feature NOUN bracis-28396 187 13 subset subset NOUN bracis-28396 187 14 performed perform VERB bracis-28396 187 15 by by ADP bracis-28396 187 16 pca pca PROPN bracis-28396 187 17 , , PUNCT bracis-28396 187 18 while while SCONJ bracis-28396 187 19 m5 m5 NOUN bracis-28396 187 20 combines combine VERB bracis-28396 187 21 mlp mlp NOUN bracis-28396 187 22 hyperparameters hyperparameter NOUN bracis-28396 187 23 tuning tune VERB bracis-28396 187 24 using use VERB bracis-28396 187 25 pso pso NOUN bracis-28396 187 26 with with ADP bracis-28396 187 27 pca pca PROPN bracis-28396 187 28 . . PUNCT bracis-28396 188 1 finally finally ADV bracis-28396 188 2 , , PUNCT bracis-28396 188 3 rs rs ADV bracis-28396 188 4 represents represent VERB bracis-28396 188 5 the the DET bracis-28396 188 6 analysis analysis NOUN bracis-28396 188 7 of of ADP bracis-28396 188 8 random random ADJ bracis-28396 188 9 combinations combination NOUN bracis-28396 188 10 of of ADP bracis-28396 188 11 mlp mlp PROPN bracis-28396 188 12 hyperparameter hyperparameter NOUN bracis-28396 188 13 values value NOUN bracis-28396 188 14 and and CCONJ bracis-28396 188 15 selected select VERB bracis-28396 188 16 features feature NOUN bracis-28396 188 17 . . PUNCT bracis-28396 189 1 the the DET bracis-28396 189 2 experiments experiment NOUN bracis-28396 189 3 carried carry VERB bracis-28396 189 4 out out ADP bracis-28396 189 5 in in ADP bracis-28396 189 6 this this DET bracis-28396 189 7 work work NOUN bracis-28396 189 8 were be AUX bracis-28396 189 9 coded code VERB bracis-28396 189 10 using use VERB bracis-28396 189 11 pythonfootnote pythonfootnote NOUN bracis-28396 189 12 1 1 NUM bracis-28396 189 13 and and CCONJ bracis-28396 189 14 r r NOUN bracis-28396 189 15     SPACE bracis-28396 190 1 [ [ PUNCT bracis-28396 190 2 18 18 NUM bracis-28396 190 3 ] ] PUNCT bracis-28396 190 4 . . PUNCT bracis-28396 191 1 further far ADV bracis-28396 191 2 , , PUNCT bracis-28396 191 3 the the DET bracis-28396 191 4 feature feature NOUN bracis-28396 191 5 extraction extraction NOUN bracis-28396 191 6 task task NOUN bracis-28396 191 7 was be AUX bracis-28396 191 8 implemented implement VERB bracis-28396 191 9 in in ADP bracis-28396 191 10 python python NOUN bracis-28396 191 11 using use VERB bracis-28396 191 12 the the DET bracis-28396 191 13 scikit scikit NOUN bracis-28396 191 14 - - PUNCT bracis-28396 191 15 imagefootnote imagefootnote ADJ bracis-28396 191 16 2 2 NUM bracis-28396 191 17 package package NOUN bracis-28396 191 18 , , PUNCT bracis-28396 191 19 while while SCONJ bracis-28396 191 20 the the DET bracis-28396 191 21 mlp mlp NOUN bracis-28396 191 22 network network NOUN bracis-28396 191 23 was be AUX bracis-28396 191 24 developed develop VERB bracis-28396 191 25 using use VERB bracis-28396 191 26 the the DET bracis-28396 191 27 rweka rweka ADJ bracis-28396 191 28 package package NOUN bracis-28396 191 29 in in ADP bracis-28396 191 30 r r NOUN bracis-28396 191 31 , , PUNCT bracis-28396 191 32 which which PRON bracis-28396 191 33 is be AUX bracis-28396 191 34 an an DET bracis-28396 191 35 interface interface NOUN bracis-28396 191 36 to to ADP bracis-28396 191 37 weka weka PROPN bracis-28396 191 38 . . PUNCT bracis-28396 192 1 finally finally ADV bracis-28396 192 2 , , PUNCT bracis-28396 192 3 pso pso PROPN bracis-28396 192 4 was be AUX bracis-28396 192 5 also also ADV bracis-28396 192 6 implemented implement VERB bracis-28396 192 7 in in ADP bracis-28396 192 8 r. r. PROPN bracis-28396 192 9 5 5 NUM bracis-28396 192 10 experimental experimental ADJ bracis-28396 192 11 results result NOUN bracis-28396 192 12 this this DET bracis-28396 192 13 section section NOUN bracis-28396 192 14 presents present VERB bracis-28396 192 15 the the DET bracis-28396 192 16 predictive predictive ADJ bracis-28396 192 17 performance performance NOUN bracis-28396 192 18 of of ADP bracis-28396 192 19 the the DET bracis-28396 192 20 anns anns NOUN bracis-28396 192 21 assessed assess VERB bracis-28396 192 22 for for ADP bracis-28396 192 23 the the DET bracis-28396 192 24 hp hp PROPN bracis-28396 192 25 - - PUNCT bracis-28396 192 26 fs fs ADJ bracis-28396 192 27 - - PUNCT bracis-28396 192 28 pso pso NOUN bracis-28396 192 29 method method NOUN bracis-28396 192 30 , , PUNCT bracis-28396 192 31 as as ADV bracis-28396 192 32 well well ADV bracis-28396 192 33 as as ADP bracis-28396 192 34 the the DET bracis-28396 192 35 baselines baseline NOUN bracis-28396 192 36 techniques technique NOUN bracis-28396 192 37 . . PUNCT bracis-28396 193 1 notice notice VERB bracis-28396 193 2 the the DET bracis-28396 193 3 results result NOUN bracis-28396 193 4 are be AUX bracis-28396 193 5 also also ADV bracis-28396 193 6 provided provide VERB bracis-28396 193 7 with with ADP bracis-28396 193 8 the the DET bracis-28396 193 9 p p NOUN bracis-28396 193 10 - - PUNCT bracis-28396 193 11 values value NOUN bracis-28396 193 12 considering consider VERB bracis-28396 193 13 the the DET bracis-28396 193 14 wilcoxon wilcoxon ADJ bracis-28396 193 15 signed sign VERB bracis-28396 193 16 - - PUNCT bracis-28396 193 17 rank rank NOUN bracis-28396 193 18 test test NOUN bracis-28396 193 19 compared compare VERB bracis-28396 193 20 to to ADP bracis-28396 193 21 the the DET bracis-28396 193 22 hp hp PROPN bracis-28396 193 23 - - PUNCT bracis-28396 193 24 fs fs ADJ bracis-28396 193 25 - - PUNCT bracis-28396 193 26 pso pso NOUN bracis-28396 193 27 method method NOUN bracis-28396 193 28 as as ADP bracis-28396 193 29 the the DET bracis-28396 193 30 reference reference NOUN bracis-28396 193 31 for for ADP bracis-28396 193 32 statistical statistical ADJ bracis-28396 193 33 analysis analysis NOUN bracis-28396 193 34 purposes purpose NOUN bracis-28396 193 35 . . PUNCT bracis-28396 194 1 further far ADV bracis-28396 194 2 , , PUNCT bracis-28396 194 3 values value NOUN bracis-28396 194 4 presented present VERB bracis-28396 194 5 in in ADP bracis-28396 194 6 bold bold ADJ bracis-28396 194 7 stand stand NOUN bracis-28396 194 8 for for ADP bracis-28396 194 9 the the DET bracis-28396 194 10 most most ADV bracis-28396 194 11 accurate accurate ADJ bracis-28396 194 12 result result NOUN bracis-28396 194 13 overall overall ADJ bracis-28396 194 14 . . PUNCT bracis-28396 195 1 5.1 5.1 NUM bracis-28396 195 2 optimization optimization NOUN bracis-28396 195 3 evaluation evaluation NOUN bracis-28396 195 4 regarding regard VERB bracis-28396 195 5 the the DET bracis-28396 195 6 optimization optimization NOUN bracis-28396 195 7 performance performance NOUN bracis-28396 195 8 over over ADP bracis-28396 195 9 the the DET bracis-28396 195 10 validation validation NOUN bracis-28396 195 11 set set NOUN bracis-28396 195 12 , , PUNCT bracis-28396 195 13 one one PRON bracis-28396 195 14 can can AUX bracis-28396 195 15 observe observe VERB bracis-28396 195 16 in in ADP bracis-28396 195 17 table table NOUN bracis-28396 195 18     SPACE bracis-28396 195 19 1 1 NUM bracis-28396 195 20 that that SCONJ bracis-28396 195 21 hp hp PROPN bracis-28396 195 22 - - PUNCT bracis-28396 195 23 fs fs NOUN bracis-28396 195 24 - - PUNCT bracis-28396 195 25 pso pso NOUN bracis-28396 195 26 and and CCONJ bracis-28396 195 27 m3 m3 PROPN bracis-28396 195 28 have have AUX bracis-28396 195 29 obtained obtain VERB bracis-28396 195 30 the the DET bracis-28396 195 31 best good ADJ bracis-28396 195 32 results result NOUN bracis-28396 195 33 , , PUNCT bracis-28396 195 34 achieving achieve VERB bracis-28396 195 35 a a DET bracis-28396 195 36 bac bac NOUN bracis-28396 195 37 average average NOUN bracis-28396 195 38 of of ADP bracis-28396 195 39 0.850 0.850 NUM bracis-28396 195 40 . . PUNCT bracis-28396 196 1 such such ADJ bracis-28396 196 2 techniques technique NOUN bracis-28396 196 3 provided provide VERB bracis-28396 196 4 an an DET bracis-28396 196 5 improvement improvement NOUN bracis-28396 196 6 of of ADP bracis-28396 196 7 around around ADV bracis-28396 196 8 \(10\%\ \(10\%\ NUM bracis-28396 196 9 ) ) PUNCT bracis-28396 196 10 compared compare VERB bracis-28396 196 11 to to ADP bracis-28396 196 12 m1 m1 PROPN bracis-28396 196 13 , , PUNCT bracis-28396 196 14 which which PRON bracis-28396 196 15 represents represent VERB bracis-28396 196 16 an an DET bracis-28396 196 17 ann ann PROPN bracis-28396 196 18 using use VERB bracis-28396 196 19 default default NOUN bracis-28396 196 20 hp hp ADJ bracis-28396 196 21 values value NOUN bracis-28396 196 22 and and CCONJ bracis-28396 196 23 the the DET bracis-28396 196 24 whole whole ADJ bracis-28396 196 25 set set NOUN bracis-28396 196 26 of of ADP bracis-28396 196 27 features feature NOUN bracis-28396 196 28 . . PUNCT bracis-28396 197 1 therefore therefore ADV bracis-28396 197 2 , , PUNCT bracis-28396 197 3 the the DET bracis-28396 197 4 most most ADV bracis-28396 197 5 important important ADJ bracis-28396 197 6 finding finding NOUN bracis-28396 197 7 from from ADP bracis-28396 197 8 these these DET bracis-28396 197 9 results result NOUN bracis-28396 197 10 is be AUX bracis-28396 197 11 that that SCONJ bracis-28396 197 12 the the DET bracis-28396 197 13 optimization optimization NOUN bracis-28396 197 14 of of ADP bracis-28396 197 15 mlp mlp PROPN bracis-28396 197 16 hyperparameters hyperparameter NOUN bracis-28396 197 17 , , PUNCT bracis-28396 197 18 and and CCONJ bracis-28396 197 19 even even ADV bracis-28396 197 20 more more ADV bracis-28396 197 21 a a DET bracis-28396 197 22 proper proper ADJ bracis-28396 197 23 selection selection NOUN bracis-28396 197 24 of of ADP bracis-28396 197 25 the the DET bracis-28396 197 26 more more ADV bracis-28396 197 27 suitable suitable ADJ bracis-28396 197 28 features feature NOUN bracis-28396 197 29 , , PUNCT bracis-28396 197 30 have have VERB bracis-28396 197 31 a a DET bracis-28396 197 32 strong strong ADJ bracis-28396 197 33 influence influence NOUN bracis-28396 197 34 in in ADP bracis-28396 197 35 the the DET bracis-28396 197 36 induction induction NOUN bracis-28396 197 37 of of ADP bracis-28396 197 38 mlp mlp PROPN bracis-28396 197 39 models model NOUN bracis-28396 197 40 applied apply VERB bracis-28396 197 41 to to ADP bracis-28396 197 42 wood wood NOUN bracis-28396 197 43 boards board NOUN bracis-28396 197 44 quality quality NOUN bracis-28396 197 45 classification classification NOUN bracis-28396 197 46 . . PUNCT bracis-28396 198 1 on on ADP bracis-28396 198 2 the the DET bracis-28396 198 3 other other ADJ bracis-28396 198 4 hand hand NOUN bracis-28396 198 5 , , PUNCT bracis-28396 198 6 results result NOUN bracis-28396 198 7 also also ADV bracis-28396 198 8 show show VERB bracis-28396 198 9 that that SCONJ bracis-28396 198 10 performing perform VERB bracis-28396 198 11 only only ADV bracis-28396 198 12 one one NUM bracis-28396 198 13 task task NOUN bracis-28396 198 14 may may AUX bracis-28396 198 15 also also ADV bracis-28396 198 16 be be AUX bracis-28396 198 17 enough enough ADJ bracis-28396 198 18 to to PART bracis-28396 198 19 increase increase VERB bracis-28396 198 20 performance performance NOUN bracis-28396 198 21 since since SCONJ bracis-28396 198 22 m3 m3 PROPN bracis-28396 198 23 performed perform VERB bracis-28396 198 24 only only ADV bracis-28396 198 25 the the DET bracis-28396 198 26 task task NOUN bracis-28396 198 27 of of ADP bracis-28396 198 28 feature feature NOUN bracis-28396 198 29 selection selection NOUN bracis-28396 198 30 , , PUNCT bracis-28396 198 31 while while SCONJ bracis-28396 198 32 m2 m2 PROPN bracis-28396 198 33 , , PUNCT bracis-28396 198 34 which which PRON bracis-28396 198 35 obtained obtain VERB bracis-28396 198 36 an an DET bracis-28396 198 37 average average ADJ bracis-28396 198 38 bac bac NOUN bracis-28396 198 39 of of ADP bracis-28396 198 40 0.834 0.834 NUM bracis-28396 198 41 , , PUNCT bracis-28396 198 42 performing perform VERB bracis-28396 198 43 only only ADV bracis-28396 198 44 the the DET bracis-28396 198 45 task task NOUN bracis-28396 198 46 of of ADP bracis-28396 198 47 mlp mlp PROPN bracis-28396 198 48 hyperparameter hyperparameter NOUN bracis-28396 198 49 tuning tune VERB bracis-28396 198 50 . . PUNCT bracis-28396 199 1 although although SCONJ bracis-28396 199 2 by by ADP bracis-28396 199 3 a a DET bracis-28396 199 4 narrow narrow ADJ bracis-28396 199 5 margin margin NOUN bracis-28396 199 6 , , PUNCT bracis-28396 199 7 in in ADP bracis-28396 199 8 this this DET bracis-28396 199 9 case case NOUN bracis-28396 199 10 , , PUNCT bracis-28396 199 11 the the DET bracis-28396 199 12 set set NOUN bracis-28396 199 13 of of ADP bracis-28396 199 14 features feature NOUN bracis-28396 199 15 was be AUX bracis-28396 199 16 more more ADV bracis-28396 199 17 expressive expressive ADJ bracis-28396 199 18 for for ADP bracis-28396 199 19 the the DET bracis-28396 199 20 model model NOUN bracis-28396 199 21 predictive predictive ADJ bracis-28396 199 22 performance performance NOUN bracis-28396 199 23 than than ADP bracis-28396 199 24 tuning tune VERB bracis-28396 199 25 the the DET bracis-28396 199 26 network network NOUN bracis-28396 199 27 hyperparameters hyperparameter NOUN bracis-28396 199 28 . . PUNCT bracis-28396 200 1 this this DET bracis-28396 200 2 behavior behavior NOUN bracis-28396 200 3 suggests suggest VERB bracis-28396 200 4 that that SCONJ bracis-28396 200 5 a a DET bracis-28396 200 6 user user NOUN bracis-28396 200 7 should should AUX bracis-28396 200 8 , , PUNCT bracis-28396 200 9 at at ADP bracis-28396 200 10 least least ADJ bracis-28396 200 11 , , PUNCT bracis-28396 200 12 select select VERB bracis-28396 200 13 a a DET bracis-28396 200 14 set set NOUN bracis-28396 200 15 of of ADP bracis-28396 200 16 features feature NOUN bracis-28396 200 17 for for ADP bracis-28396 200 18 the the DET bracis-28396 200 19 hyperparameter hyperparameter NOUN bracis-28396 200 20 values value NOUN bracis-28396 200 21 defined define VERB bracis-28396 200 22 a a DET bracis-28396 200 23 priori priori ADV bracis-28396 200 24 . . PUNCT bracis-28396 201 1 besides besides SCONJ bracis-28396 201 2 the the DET bracis-28396 201 3 bac bac PROPN bracis-28396 201 4 values value NOUN bracis-28396 201 5 , , PUNCT bracis-28396 201 6 table table NOUN bracis-28396 201 7     SPACE bracis-28396 201 8 1 1 NUM bracis-28396 201 9 also also ADV bracis-28396 201 10 provides provide VERB bracis-28396 201 11 the the DET bracis-28396 201 12 p p NOUN bracis-28396 201 13 - - PUNCT bracis-28396 201 14 values value NOUN bracis-28396 201 15 considering consider VERB bracis-28396 201 16 the the DET bracis-28396 201 17 wilcoxon wilcoxon ADJ bracis-28396 201 18 signed sign VERB bracis-28396 201 19 - - PUNCT bracis-28396 201 20 rank rank NOUN bracis-28396 201 21 test test NOUN bracis-28396 201 22 compared compare VERB bracis-28396 201 23 to to ADP bracis-28396 201 24 the the DET bracis-28396 201 25 hp hp PROPN bracis-28396 201 26 - - PUNCT bracis-28396 201 27 fs fs ADJ bracis-28396 201 28 - - PUNCT bracis-28396 201 29 pso pso NOUN bracis-28396 201 30 method method NOUN bracis-28396 201 31 as as ADP bracis-28396 201 32 the the DET bracis-28396 201 33 reference reference NOUN bracis-28396 201 34 . . PUNCT bracis-28396 202 1 these these DET bracis-28396 202 2 p p NOUN bracis-28396 202 3 - - PUNCT bracis-28396 202 4 values value NOUN bracis-28396 202 5 support support VERB bracis-28396 202 6 our our PRON bracis-28396 202 7 previous previous ADJ bracis-28396 202 8 observations observation NOUN bracis-28396 202 9 , , PUNCT bracis-28396 202 10 considering consider VERB bracis-28396 202 11 only only ADV bracis-28396 202 12 m3 m3 PROPN bracis-28396 202 13 and and CCONJ bracis-28396 202 14 the the DET bracis-28396 202 15 random random ADJ bracis-28396 202 16 search search NOUN bracis-28396 202 17 obtained obtain VERB bracis-28396 202 18 a a DET bracis-28396 202 19 significance significance NOUN bracis-28396 202 20 level level NOUN bracis-28396 202 21 higher high ADJ bracis-28396 202 22 than than ADP bracis-28396 202 23 \(\alpha \(\alpha NOUN bracis-28396 202 24 = = ADJ bracis-28396 202 25 0.05\ 0.05\ NUM bracis-28396 202 26 ) ) PUNCT bracis-28396 202 27 . . PUNCT bracis-28396 203 1 table table NOUN bracis-28396 203 2 1 1 NUM bracis-28396 203 3 . . PUNCT bracis-28396 203 4 average average ADJ bracis-28396 203 5 bac bac NOUN bracis-28396 203 6 and and CCONJ bracis-28396 203 7 the the DET bracis-28396 203 8 standard standard ADJ bracis-28396 203 9 deviation deviation NOUN bracis-28396 203 10 concerning concern VERB bracis-28396 203 11 the the DET bracis-28396 203 12 task task NOUN bracis-28396 203 13 of of ADP bracis-28396 203 14 mlp mlp PROPN bracis-28396 203 15 hyperparameter hyperparameter NOUN bracis-28396 203 16 tuning tuning NOUN bracis-28396 203 17 and and CCONJ bracis-28396 203 18 feature feature NOUN bracis-28396 203 19 selection selection NOUN bracis-28396 203 20 , , PUNCT bracis-28396 203 21 considering consider VERB bracis-28396 203 22 the the DET bracis-28396 203 23 validation validation NOUN bracis-28396 203 24 set set VERB bracis-28396 203 25 over over ADP bracis-28396 203 26 10 10 NUM bracis-28396 203 27 executions execution NOUN bracis-28396 203 28 . . PUNCT bracis-28396 204 1 notice notice VERB bracis-28396 204 2 the the DET bracis-28396 204 3 “ " PUNCT bracis-28396 204 4 p p NOUN bracis-28396 204 5 - - PUNCT bracis-28396 204 6 values value NOUN bracis-28396 204 7 ” " PUNCT bracis-28396 204 8 are be AUX bracis-28396 204 9 compared compare VERB bracis-28396 204 10 against against ADP bracis-28396 204 11 the the DET bracis-28396 204 12 hp hp PROPN bracis-28396 204 13 - - PUNCT bracis-28396 204 14 fs fs ADJ bracis-28396 204 15 - - PUNCT bracis-28396 204 16 pso pso NOUN bracis-28396 204 17 reference.full reference.full NOUN bracis-28396 204 18 size size NOUN bracis-28396 204 19 table table NOUN bracis-28396 204 20 notice notice VERB bracis-28396 204 21 the the DET bracis-28396 204 22 positive positive ADJ bracis-28396 204 23 behavior behavior NOUN bracis-28396 204 24 of of ADP bracis-28396 204 25 the the DET bracis-28396 204 26 random random ADJ bracis-28396 204 27 search search NOUN bracis-28396 204 28 approach approach NOUN bracis-28396 204 29 , , PUNCT bracis-28396 204 30 which which PRON bracis-28396 204 31 is be AUX bracis-28396 204 32 somehow somehow ADV bracis-28396 204 33 expected expect VERB bracis-28396 204 34 since since SCONJ bracis-28396 204 35 the the DET bracis-28396 204 36 model model NOUN bracis-28396 204 37 presents present VERB bracis-28396 204 38 itself itself PRON bracis-28396 204 39 as as ADP bracis-28396 204 40 more more ADV bracis-28396 204 41 sensitive sensitive ADJ bracis-28396 204 42 to to ADP bracis-28396 204 43 a a DET bracis-28396 204 44 proper proper ADJ bracis-28396 204 45 selection selection NOUN bracis-28396 204 46 of of ADP bracis-28396 204 47 the the DET bracis-28396 204 48 features feature NOUN bracis-28396 204 49 instead instead ADV bracis-28396 204 50 of of ADP bracis-28396 204 51 the the DET bracis-28396 204 52 network network NOUN bracis-28396 204 53 hyperparameter hyperparameter NOUN bracis-28396 204 54 tuning tuning NOUN bracis-28396 204 55 , , PUNCT bracis-28396 204 56 which which PRON bracis-28396 204 57 is be AUX bracis-28396 204 58 expected expect VERB bracis-28396 204 59 to to PART bracis-28396 204 60 be be AUX bracis-28396 204 61 a a DET bracis-28396 204 62 more more ADV bracis-28396 204 63 straightforward straightforward ADJ bracis-28396 204 64 task task NOUN bracis-28396 204 65 due due ADP bracis-28396 204 66 to to ADP bracis-28396 204 67 the the DET bracis-28396 204 68 binary binary ADJ bracis-28396 204 69 nature nature NOUN bracis-28396 204 70 of of ADP bracis-28396 204 71 the the DET bracis-28396 204 72 search search NOUN bracis-28396 204 73 space space NOUN bracis-28396 204 74 . . PUNCT bracis-28396 205 1 further far ADV bracis-28396 205 2 , , PUNCT bracis-28396 205 3 the the DET bracis-28396 205 4 data datum NOUN bracis-28396 205 5 dimensionality dimensionality NOUN bracis-28396 205 6 reduction reduction NOUN bracis-28396 205 7 performed perform VERB bracis-28396 205 8 by by ADP bracis-28396 205 9 pca pca PROPN bracis-28396 205 10 showed show VERB bracis-28396 205 11 to to PART bracis-28396 205 12 be be AUX bracis-28396 205 13 inadequate inadequate ADJ bracis-28396 205 14 for for ADP bracis-28396 205 15 this this DET bracis-28396 205 16 problem problem NOUN bracis-28396 205 17 , , PUNCT bracis-28396 205 18 as as SCONJ bracis-28396 205 19 denoted denote VERB bracis-28396 205 20 by by ADP bracis-28396 205 21 methods method NOUN bracis-28396 205 22 m4 m4 PROPN bracis-28396 205 23 , , PUNCT bracis-28396 205 24 which which PRON bracis-28396 205 25 obtained obtain VERB bracis-28396 205 26 bac bac PROPN bracis-28396 205 27 average average ADJ bracis-28396 205 28 results result NOUN bracis-28396 205 29 lower low ADJ bracis-28396 205 30 than than ADP bracis-28396 205 31 using use VERB bracis-28396 205 32 a a DET bracis-28396 205 33 default default NOUN bracis-28396 205 34 configuration configuration NOUN bracis-28396 205 35 , , PUNCT bracis-28396 205 36 and and CCONJ bracis-28396 205 37 m5 m5 NOUN bracis-28396 205 38 . . PUNCT bracis-28396 206 1 the the DET bracis-28396 206 2 main main ADJ bracis-28396 206 3 reason reason NOUN bracis-28396 206 4 lies lie VERB bracis-28396 206 5 in in ADP bracis-28396 206 6 the the DET bracis-28396 206 7 fact fact NOUN bracis-28396 206 8 that that SCONJ bracis-28396 206 9 pca pca PROPN bracis-28396 206 10 may may AUX bracis-28396 206 11 not not PART bracis-28396 206 12 be be AUX bracis-28396 206 13 able able ADJ bracis-28396 206 14 to to PART bracis-28396 206 15 describe describe VERB bracis-28396 206 16 sufficiently sufficiently ADV bracis-28396 206 17 well well ADV bracis-28396 206 18 the the DET bracis-28396 206 19 problem problem NOUN bracis-28396 206 20 due due ADP bracis-28396 206 21 to to ADP bracis-28396 206 22 its its PRON bracis-28396 206 23 linear linear ADJ bracis-28396 206 24 nature nature NOUN bracis-28396 206 25 . . PUNCT bracis-28396 207 1 for for ADP bracis-28396 207 2 a a DET bracis-28396 207 3 better well ADJ bracis-28396 207 4 understanding understanding NOUN bracis-28396 207 5 of of ADP bracis-28396 207 6 each each DET bracis-28396 207 7 method method NOUN bracis-28396 207 8 ’s ’s PART bracis-28396 207 9 behavior behavior NOUN bracis-28396 207 10 during during ADP bracis-28396 207 11 pso pso NOUN bracis-28396 207 12 convergence convergence NOUN bracis-28396 207 13 , , PUNCT bracis-28396 207 14 fig fig NOUN bracis-28396 207 15 . . PUNCT bracis-28396 207 16     SPACE bracis-28396 207 17 3 3 NUM bracis-28396 207 18 depicts depict VERB bracis-28396 207 19 the the DET bracis-28396 207 20 optimization optimization NOUN bracis-28396 207 21 performance performance NOUN bracis-28396 207 22 ( ( PUNCT bracis-28396 207 23 a a X bracis-28396 207 24 ) ) PUNCT bracis-28396 207 25 and and CCONJ bracis-28396 207 26 the the DET bracis-28396 207 27 evolution evolution NOUN bracis-28396 207 28 of of ADP bracis-28396 207 29 the the DET bracis-28396 207 30 bac bac PROPN bracis-28396 207 31 values value NOUN bracis-28396 207 32 ( ( PUNCT bracis-28396 207 33 b b NOUN bracis-28396 207 34 ) ) PUNCT bracis-28396 207 35 during during ADP bracis-28396 207 36 300 300 NUM bracis-28396 207 37 iterations iteration NOUN bracis-28396 207 38 . . PUNCT bracis-28396 208 1 figure figure NOUN bracis-28396 208 2     SPACE bracis-28396 208 3 3 3 NUM bracis-28396 208 4 ( ( PUNCT bracis-28396 208 5 a a NOUN bracis-28396 208 6 ) ) PUNCT bracis-28396 208 7 considers consider VERB bracis-28396 208 8 the the DET bracis-28396 208 9 average average ADJ bracis-28396 208 10 values value NOUN bracis-28396 208 11 over over ADP bracis-28396 208 12 10 10 NUM bracis-28396 208 13 runs run NOUN bracis-28396 208 14 , , PUNCT bracis-28396 208 15 where where SCONJ bracis-28396 208 16 each each DET bracis-28396 208 17 iteration iteration NOUN bracis-28396 208 18 in in ADP bracis-28396 208 19 the the DET bracis-28396 208 20 rs rs ADJ bracis-28396 208 21 curve curve NOUN bracis-28396 208 22 reflects reflect VERB bracis-28396 208 23 the the DET bracis-28396 208 24 average average ADJ bracis-28396 208 25 evaluation evaluation NOUN bracis-28396 208 26 among among ADP bracis-28396 208 27 30 30 NUM bracis-28396 208 28 executions execution NOUN bracis-28396 208 29 , , PUNCT bracis-28396 208 30 i.e. i.e. X bracis-28396 208 31 , , PUNCT bracis-28396 208 32 the the DET bracis-28396 208 33 same same ADJ bracis-28396 208 34 number number NOUN bracis-28396 208 35 of of ADP bracis-28396 208 36 assessments assessment NOUN bracis-28396 208 37 performed perform VERB bracis-28396 208 38 by by ADP bracis-28396 208 39 pso pso NOUN bracis-28396 208 40 considering consider VERB bracis-28396 208 41 30 30 NUM bracis-28396 208 42 particles particle NOUN bracis-28396 208 43 . . PUNCT bracis-28396 209 1 finally finally ADV bracis-28396 209 2 , , PUNCT bracis-28396 209 3 m1 m1 PROPN bracis-28396 209 4 and and CCONJ bracis-28396 209 5 m4 m4 PROPN bracis-28396 209 6 are be AUX bracis-28396 209 7 represented represent VERB bracis-28396 209 8 by by ADP bracis-28396 209 9 fixed fix VERB bracis-28396 209 10 lines line NOUN bracis-28396 209 11 since since SCONJ bracis-28396 209 12 no no DET bracis-28396 209 13 optimization optimization NOUN bracis-28396 209 14 was be AUX bracis-28396 209 15 performed perform VERB bracis-28396 209 16 over over ADP bracis-28396 209 17 such such ADJ bracis-28396 209 18 approaches approach NOUN bracis-28396 209 19 . . PUNCT bracis-28396 210 1 figure figure NOUN bracis-28396 210 2     SPACE bracis-28396 210 3 3(b 3(b NUM bracis-28396 210 4 ) ) PUNCT bracis-28396 210 5 corroborates corroborate VERB bracis-28396 210 6 our our PRON bracis-28396 210 7 claim claim NOUN bracis-28396 210 8 that that SCONJ bracis-28396 210 9 hp hp PROPN bracis-28396 210 10 - - PUNCT bracis-28396 210 11 fs fs NOUN bracis-28396 210 12 - - PUNCT bracis-28396 210 13 pso pso NOUN bracis-28396 210 14 performed perform VERB bracis-28396 210 15 a a DET bracis-28396 210 16 guided guide VERB bracis-28396 210 17 search search NOUN bracis-28396 210 18 through through ADP bracis-28396 210 19 the the DET bracis-28396 210 20 mlp mlp NOUN bracis-28396 210 21 hyperparameters hyperparameter NOUN bracis-28396 210 22 and and CCONJ bracis-28396 210 23 features feature NOUN bracis-28396 210 24 spaces space NOUN bracis-28396 210 25 , , PUNCT bracis-28396 210 26 improving improve VERB bracis-28396 210 27 its its PRON bracis-28396 210 28 performance performance NOUN bracis-28396 210 29 over over ADP bracis-28396 210 30 the the DET bracis-28396 210 31 iterations iteration NOUN bracis-28396 210 32 . . PUNCT bracis-28396 211 1 notice notice NOUN bracis-28396 211 2 pso pso NOUN bracis-28396 211 3 can can AUX bracis-28396 211 4 reduce reduce VERB bracis-28396 211 5 the the DET bracis-28396 211 6 number number NOUN bracis-28396 211 7 of of ADP bracis-28396 211 8 iterations iteration NOUN bracis-28396 211 9 required require VERB bracis-28396 211 10 for for ADP bracis-28396 211 11 finding find VERB bracis-28396 211 12 reasonable reasonable ADJ bracis-28396 211 13 bac bac NOUN bracis-28396 211 14 values value NOUN bracis-28396 211 15 since since SCONJ bracis-28396 211 16 it it PRON bracis-28396 211 17 obtained obtain VERB bracis-28396 211 18 relatively relatively ADV bracis-28396 211 19 high high ADJ bracis-28396 211 20 accuracies accuracy NOUN bracis-28396 211 21 ( ( PUNCT bracis-28396 211 22 around around ADP bracis-28396 211 23 0.840 0.840 NUM bracis-28396 211 24 ) ) PUNCT bracis-28396 211 25 after after ADP bracis-28396 211 26 80 80 NUM bracis-28396 211 27 iterations iteration NOUN bracis-28396 211 28 only only ADV bracis-28396 211 29 . . PUNCT bracis-28396 212 1 fig fig NOUN bracis-28396 212 2 . . PUNCT bracis-28396 213 1 3 3 X bracis-28396 213 2 . . X bracis-28396 213 3 pso pso NOUN bracis-28396 213 4 convergence convergence NOUN bracis-28396 213 5 considering consider VERB bracis-28396 213 6 the the DET bracis-28396 213 7 evaluation evaluation NOUN bracis-28396 213 8 dataset dataset NOUN bracis-28396 213 9 ( ( PUNCT bracis-28396 213 10 a a NOUN bracis-28396 213 11 ) ) PUNCT bracis-28396 213 12 and and CCONJ bracis-28396 213 13 evolution evolution NOUN bracis-28396 213 14 of of ADP bracis-28396 213 15 the the DET bracis-28396 213 16 bac bac PROPN bracis-28396 213 17 values value NOUN bracis-28396 213 18 of of ADP bracis-28396 213 19 hp hp PROPN bracis-28396 213 20 - - PUNCT bracis-28396 213 21 fs fs NOUN bracis-28396 213 22 - - PUNCT bracis-28396 213 23 pso pso NOUN bracis-28396 213 24 and and CCONJ bracis-28396 213 25 rs rs NOUN bracis-28396 213 26 . . PUNCT bracis-28396 214 1 in in ADP bracis-28396 214 2 this this DET bracis-28396 214 3 case case NOUN bracis-28396 214 4 , , PUNCT bracis-28396 214 5 the the DET bracis-28396 214 6 best good ADJ bracis-28396 214 7 bac bac NOUN bracis-28396 214 8 value value NOUN bracis-28396 214 9 found find VERB bracis-28396 214 10 by by ADP bracis-28396 214 11 rs rs PROPN bracis-28396 214 12 up up ADP bracis-28396 214 13 to to ADP bracis-28396 214 14 a a DET bracis-28396 214 15 iteration iteration NOUN bracis-28396 214 16 is be AUX bracis-28396 214 17 kept keep VERB bracis-28396 214 18 in in ADP bracis-28396 214 19 the the DET bracis-28396 214 20 next next ADJ bracis-28396 214 21 iterations iteration NOUN bracis-28396 214 22 ( ( PUNCT bracis-28396 214 23 b b NOUN bracis-28396 214 24 ) ) PUNCT bracis-28396 214 25 . . PUNCT bracis-28396 215 1 full full ADJ bracis-28396 215 2 size size NOUN bracis-28396 215 3 image image NOUN bracis-28396 215 4 besides besides SCONJ bracis-28396 215 5 , , PUNCT bracis-28396 215 6 fig fig NOUN bracis-28396 215 7 . . PUNCT bracis-28396 215 8     SPACE bracis-28396 215 9 3(b 3(b NUM bracis-28396 215 10 ) ) PUNCT bracis-28396 215 11 depicts depict VERB bracis-28396 215 12 the the DET bracis-28396 215 13 hp hp PROPN bracis-28396 215 14 - - PUNCT bracis-28396 215 15 fs fs ADJ bracis-28396 215 16 - - PUNCT bracis-28396 215 17 pso pso NOUN bracis-28396 215 18 performance performance NOUN bracis-28396 215 19 compared compare VERB bracis-28396 215 20 against against ADP bracis-28396 215 21 a a DET bracis-28396 215 22 random random ADJ bracis-28396 215 23 search search NOUN bracis-28396 215 24 considering consider VERB bracis-28396 215 25 the the DET bracis-28396 215 26 best good ADJ bracis-28396 215 27 results result NOUN bracis-28396 215 28 over over ADP bracis-28396 215 29 each each DET bracis-28396 215 30 iteration iteration NOUN bracis-28396 215 31 , , PUNCT bracis-28396 215 32 instead instead ADV bracis-28396 215 33 of of ADP bracis-28396 215 34 an an DET bracis-28396 215 35 average average NOUN bracis-28396 215 36 . . PUNCT bracis-28396 216 1 one one PRON bracis-28396 216 2 can can AUX bracis-28396 216 3 observe observe VERB bracis-28396 216 4 the the DET bracis-28396 216 5 random random ADJ bracis-28396 216 6 search search NOUN bracis-28396 216 7 performed perform VERB bracis-28396 216 8 slightly slightly ADV bracis-28396 216 9 better well ADV bracis-28396 216 10 during during ADP bracis-28396 216 11 the the DET bracis-28396 216 12 first first ADJ bracis-28396 216 13 30 30 NUM bracis-28396 216 14 iterations iteration NOUN bracis-28396 216 15 . . PUNCT bracis-28396 217 1 afterward afterward ADV bracis-28396 217 2 , , PUNCT bracis-28396 217 3 the the DET bracis-28396 217 4 hp hp PROPN bracis-28396 217 5 - - PUNCT bracis-28396 217 6 fs fs NOUN bracis-28396 217 7 - - PUNCT bracis-28396 217 8 pso pso NOUN bracis-28396 217 9 surpassed surpass VERB bracis-28396 217 10 rs rs ADV bracis-28396 217 11 and and CCONJ bracis-28396 217 12 kept keep VERB bracis-28396 217 13 this this DET bracis-28396 217 14 advantage advantage NOUN bracis-28396 217 15 until until ADP bracis-28396 217 16 reaching reach VERB bracis-28396 217 17 the the DET bracis-28396 217 18 300 300 NUM bracis-28396 217 19 iterations iteration NOUN bracis-28396 217 20 . . PUNCT bracis-28396 218 1 as as SCONJ bracis-28396 218 2 previously previously ADV bracis-28396 218 3 mentioned mention VERB bracis-28396 218 4 , , PUNCT bracis-28396 218 5 it it PRON bracis-28396 218 6 is be AUX bracis-28396 218 7 possible possible ADJ bracis-28396 218 8 to to PART bracis-28396 218 9 note note VERB bracis-28396 218 10 that that SCONJ bracis-28396 218 11 there there PRON bracis-28396 218 12 is be VERB bracis-28396 218 13 a a DET bracis-28396 218 14 considerable considerable ADJ bracis-28396 218 15 improvement improvement NOUN bracis-28396 218 16 of of ADP bracis-28396 218 17 bac bac NOUN bracis-28396 218 18 for for ADP bracis-28396 218 19 both both DET bracis-28396 218 20 methods method NOUN bracis-28396 218 21 in in ADP bracis-28396 218 22 the the DET bracis-28396 218 23 first first ADJ bracis-28396 218 24 100 100 NUM bracis-28396 218 25 iterations iteration NOUN bracis-28396 218 26 , , PUNCT bracis-28396 218 27 and and CCONJ bracis-28396 218 28 then then ADV bracis-28396 218 29 there there PRON bracis-28396 218 30 is be VERB bracis-28396 218 31 a a DET bracis-28396 218 32 slowdown slowdown NOUN bracis-28396 218 33 in in ADP bracis-28396 218 34 the the DET bracis-28396 218 35 bac bac NOUN bracis-28396 218 36 growth growth NOUN bracis-28396 218 37 . . PUNCT bracis-28396 219 1 such such DET bracis-28396 219 2 a a DET bracis-28396 219 3 piece piece NOUN bracis-28396 219 4 of of ADP bracis-28396 219 5 information information NOUN bracis-28396 219 6 is be AUX bracis-28396 219 7 of of ADP bracis-28396 219 8 extreme extreme ADJ bracis-28396 219 9 relevance relevance NOUN bracis-28396 219 10 for for ADP bracis-28396 219 11 industrial industrial ADJ bracis-28396 219 12 applications application NOUN bracis-28396 219 13 , , PUNCT bracis-28396 219 14 since since SCONJ bracis-28396 219 15 it it PRON bracis-28396 219 16 may may AUX bracis-28396 219 17 save save VERB bracis-28396 219 18 time time NOUN bracis-28396 219 19 and and CCONJ bracis-28396 219 20 effort effort NOUN bracis-28396 219 21 during during ADP bracis-28396 219 22 the the DET bracis-28396 219 23 task task NOUN bracis-28396 219 24 of of ADP bracis-28396 219 25 tuning tune VERB bracis-28396 219 26 the the DET bracis-28396 219 27 model model NOUN bracis-28396 219 28 ’s ’s PART bracis-28396 219 29 hyperparameter hyperparameter NOUN bracis-28396 219 30 selecting select VERB bracis-28396 219 31 the the DET bracis-28396 219 32 best good ADJ bracis-28396 219 33 subset subset NOUN bracis-28396 219 34 of of ADP bracis-28396 219 35 features feature NOUN bracis-28396 219 36 . . PUNCT bracis-28396 220 1 5.2 5.2 NUM bracis-28396 220 2 classification classification NOUN bracis-28396 220 3 this this DET bracis-28396 220 4 section section NOUN bracis-28396 220 5 investigates investigate VERB bracis-28396 220 6 the the DET bracis-28396 220 7 ann ann PROPN bracis-28396 220 8 generalization generalization NOUN bracis-28396 220 9 power power NOUN bracis-28396 220 10 by by ADP bracis-28396 220 11 evaluating evaluate VERB bracis-28396 220 12 the the DET bracis-28396 220 13 predictive predictive ADJ bracis-28396 220 14 performance performance NOUN bracis-28396 220 15 of of ADP bracis-28396 220 16 the the DET bracis-28396 220 17 model model NOUN bracis-28396 220 18 over over ADP bracis-28396 220 19 the the DET bracis-28396 220 20 testing testing NOUN bracis-28396 220 21 set set NOUN bracis-28396 220 22 , , PUNCT bracis-28396 220 23 considering consider VERB bracis-28396 220 24 the the DET bracis-28396 220 25 best good ADJ bracis-28396 220 26 set set NOUN bracis-28396 220 27 of of ADP bracis-28396 220 28 hyperparameters hyperparameter NOUN bracis-28396 220 29 and and CCONJ bracis-28396 220 30 a a DET bracis-28396 220 31 subset subset NOUN bracis-28396 220 32 of of ADP bracis-28396 220 33 features feature NOUN bracis-28396 220 34 found find VERB bracis-28396 220 35 during during ADP bracis-28396 220 36 the the DET bracis-28396 220 37 optimization optimization NOUN bracis-28396 220 38 process process NOUN bracis-28396 220 39 . . PUNCT bracis-28396 221 1 table table NOUN bracis-28396 221 2     SPACE bracis-28396 221 3 2 2 NUM bracis-28396 221 4 presents present VERB bracis-28396 221 5 the the DET bracis-28396 221 6 bac bac PROPN bracis-28396 221 7 values value NOUN bracis-28396 221 8 obtained obtain VERB bracis-28396 221 9 in in ADP bracis-28396 221 10 this this DET bracis-28396 221 11 context context NOUN bracis-28396 221 12 . . PUNCT bracis-28396 222 1 notice notice VERB bracis-28396 222 2 the the DET bracis-28396 222 3 values value NOUN bracis-28396 222 4 presented present VERB bracis-28396 222 5 in in ADP bracis-28396 222 6 bold bold ADJ bracis-28396 222 7 stand stand NOUN bracis-28396 222 8 for for ADP bracis-28396 222 9 the the DET bracis-28396 222 10 most most ADV bracis-28396 222 11 accurate accurate ADJ bracis-28396 222 12 approach approach NOUN bracis-28396 222 13 overall overall ADJ bracis-28396 222 14 . . PUNCT bracis-28396 223 1 table table NOUN bracis-28396 223 2 2 2 NUM bracis-28396 223 3 . . X bracis-28396 223 4 average average ADJ bracis-28396 223 5 bac bac NOUN bracis-28396 223 6 and and CCONJ bracis-28396 223 7 the the DET bracis-28396 223 8 standard standard ADJ bracis-28396 223 9 deviation deviation NOUN bracis-28396 223 10 concerning concern VERB bracis-28396 223 11 the the DET bracis-28396 223 12 task task NOUN bracis-28396 223 13 of of ADP bracis-28396 223 14 mlp mlp PROPN bracis-28396 223 15 wood wood NOUN bracis-28396 223 16 quality quality NOUN bracis-28396 223 17 classification classification NOUN bracis-28396 223 18 considered consider VERB bracis-28396 223 19 the the DET bracis-28396 223 20 testing testing NOUN bracis-28396 223 21 samples sample NOUN bracis-28396 223 22 and and CCONJ bracis-28396 223 23 the the DET bracis-28396 223 24 best good ADJ bracis-28396 223 25 set set NOUN bracis-28396 223 26 of of ADP bracis-28396 223 27 hyperparameters hyperparameter NOUN bracis-28396 223 28 and and CCONJ bracis-28396 223 29 subfeatures subfeature NOUN bracis-28396 223 30 found find VERB bracis-28396 223 31 during during ADP bracis-28396 223 32 the the DET bracis-28396 223 33 optimization optimization NOUN bracis-28396 223 34 process process NOUN bracis-28396 223 35 for for ADP bracis-28396 223 36 each each DET bracis-28396 223 37 approach approach NOUN bracis-28396 223 38 . . PUNCT bracis-28396 224 1 notice notice VERB bracis-28396 224 2 the the DET bracis-28396 224 3 “ " PUNCT bracis-28396 224 4 p p NOUN bracis-28396 224 5 - - PUNCT bracis-28396 224 6 values value NOUN bracis-28396 224 7 ” " PUNCT bracis-28396 224 8 are be AUX bracis-28396 224 9 compared compare VERB bracis-28396 224 10 against against ADP bracis-28396 224 11 the the DET bracis-28396 224 12 hp hp PROPN bracis-28396 224 13 - - PUNCT bracis-28396 224 14 fs fs ADJ bracis-28396 224 15 - - PUNCT bracis-28396 224 16 pso pso NOUN bracis-28396 224 17 reference.full reference.full NOUN bracis-28396 224 18 size size NOUN bracis-28396 224 19 table table NOUN bracis-28396 224 20 in in ADP bracis-28396 224 21 general general ADJ bracis-28396 224 22 , , PUNCT bracis-28396 224 23 the the DET bracis-28396 224 24 results result NOUN bracis-28396 224 25 of of ADP bracis-28396 224 26 test test NOUN bracis-28396 224 27 data datum NOUN bracis-28396 224 28 are be AUX bracis-28396 224 29 in in ADP bracis-28396 224 30 agreement agreement NOUN bracis-28396 224 31 with with ADP bracis-28396 224 32 those those PRON bracis-28396 224 33 of of ADP bracis-28396 224 34 the the DET bracis-28396 224 35 validation validation NOUN bracis-28396 224 36 set set NOUN bracis-28396 224 37 . . PUNCT bracis-28396 225 1 the the DET bracis-28396 225 2 baseline baseline PROPN bracis-28396 225 3 method method NOUN bracis-28396 225 4 m3 m3 PROPN bracis-28396 225 5 led lead VERB bracis-28396 225 6 to to ADP bracis-28396 225 7 the the DET bracis-28396 225 8 best good ADJ bracis-28396 225 9 bac bac NOUN bracis-28396 225 10 value value NOUN bracis-28396 225 11 , , PUNCT bracis-28396 225 12 followed follow VERB bracis-28396 225 13 closely closely ADV bracis-28396 225 14 by by ADP bracis-28396 225 15 hp hp PROPN bracis-28396 225 16 - - PUNCT bracis-28396 225 17 fs fs NOUN bracis-28396 225 18 - - PUNCT bracis-28396 225 19 pso pso NOUN bracis-28396 225 20 ( ( PUNCT bracis-28396 225 21 differing differ VERB bracis-28396 225 22 only only ADV bracis-28396 225 23 in in ADP bracis-28396 225 24 the the DET bracis-28396 225 25 third third ADJ bracis-28396 225 26 decimal decimal ADJ bracis-28396 225 27 place place NOUN bracis-28396 225 28 ) ) PUNCT bracis-28396 225 29 . . PUNCT bracis-28396 226 1 these these DET bracis-28396 226 2 performances performance NOUN bracis-28396 226 3 are be AUX bracis-28396 226 4 again again ADV bracis-28396 226 5 superior superior ADJ bracis-28396 226 6 to to ADP bracis-28396 226 7 m1 m1 PROPN bracis-28396 226 8 , , PUNCT bracis-28396 226 9 supporting support VERB bracis-28396 226 10 the the DET bracis-28396 226 11 idea idea NOUN bracis-28396 226 12 of of ADP bracis-28396 226 13 mlp mlp NOUN bracis-28396 226 14 hyperparameter hyperparameter NOUN bracis-28396 226 15 tuning tuning NOUN bracis-28396 226 16 and and CCONJ bracis-28396 226 17 feature feature NOUN bracis-28396 226 18 selection selection NOUN bracis-28396 226 19 influence influence NOUN bracis-28396 226 20 . . PUNCT bracis-28396 227 1 however however ADV bracis-28396 227 2 , , PUNCT bracis-28396 227 3 differently differently ADV bracis-28396 227 4 from from ADP bracis-28396 227 5 table table NOUN bracis-28396 227 6     SPACE bracis-28396 227 7 1 1 NUM bracis-28396 227 8 , , PUNCT bracis-28396 227 9 hp hp ADJ bracis-28396 227 10 - - PUNCT bracis-28396 227 11 fs fs NOUN bracis-28396 227 12 - - PUNCT bracis-28396 227 13 pso pso NOUN bracis-28396 227 14 was be AUX bracis-28396 227 15 not not PART bracis-28396 227 16 statistically statistically ADV bracis-28396 227 17 better well ADJ bracis-28396 227 18 than than ADP bracis-28396 227 19 m3 m3 PROPN bracis-28396 227 20 , , PUNCT bracis-28396 227 21 as as SCONJ bracis-28396 227 22 observed observe VERB bracis-28396 227 23 in in ADP bracis-28396 227 24 the the DET bracis-28396 227 25 p p NOUN bracis-28396 227 26 - - PUNCT bracis-28396 227 27 value value NOUN bracis-28396 227 28 \(>0.05\ \(>0.05\ PROPN bracis-28396 227 29 ) ) PUNCT bracis-28396 227 30 . . PUNCT bracis-28396 228 1 therefore therefore ADV bracis-28396 228 2 , , PUNCT bracis-28396 228 3 the the DET bracis-28396 228 4 improvement improvement NOUN bracis-28396 228 5 obtained obtain VERB bracis-28396 228 6 during during ADP bracis-28396 228 7 the the DET bracis-28396 228 8 validation validation NOUN bracis-28396 228 9 steps step NOUN bracis-28396 228 10 was be AUX bracis-28396 228 11 not not PART bracis-28396 228 12 enough enough ADJ bracis-28396 228 13 to to PART bracis-28396 228 14 provide provide VERB bracis-28396 228 15 a a DET bracis-28396 228 16 statistical statistical ADJ bracis-28396 228 17 difference difference NOUN bracis-28396 228 18 , , PUNCT bracis-28396 228 19 considering consider VERB bracis-28396 228 20 the the DET bracis-28396 228 21 test test NOUN bracis-28396 228 22 data datum NOUN bracis-28396 228 23 . . PUNCT bracis-28396 229 1 the the DET bracis-28396 229 2 high high ADJ bracis-28396 229 3 standard standard ADJ bracis-28396 229 4 deviation deviation NOUN bracis-28396 229 5 over over ADP bracis-28396 229 6 the the DET bracis-28396 229 7 testing testing NOUN bracis-28396 229 8 set set NOUN bracis-28396 229 9 explains explain VERB bracis-28396 229 10 such such ADJ bracis-28396 229 11 behavior behavior NOUN bracis-28396 229 12 , , PUNCT bracis-28396 229 13 which which PRON bracis-28396 229 14 was be AUX bracis-28396 229 15 considerably considerably ADV bracis-28396 229 16 smaller small ADJ bracis-28396 229 17 regarding regard VERB bracis-28396 229 18 the the DET bracis-28396 229 19 optimization optimization NOUN bracis-28396 229 20 steps step NOUN bracis-28396 229 21 . . PUNCT bracis-28396 230 1 6 6 NUM bracis-28396 230 2 conclusion conclusion NOUN bracis-28396 230 3 this this DET bracis-28396 230 4 paper paper NOUN bracis-28396 230 5 analyzed analyze VERB bracis-28396 230 6 the the DET bracis-28396 230 7 compound compound NOUN bracis-28396 230 8 problem problem NOUN bracis-28396 230 9 of of ADP bracis-28396 230 10 ann ann PROPN bracis-28396 230 11 hyperparameters hyperparameter NOUN bracis-28396 230 12 tuning tune VERB bracis-28396 230 13 and and CCONJ bracis-28396 230 14 feature feature NOUN bracis-28396 230 15 selection selection NOUN bracis-28396 230 16 for for ADP bracis-28396 230 17 the the DET bracis-28396 230 18 quality quality NOUN bracis-28396 230 19 classification classification NOUN bracis-28396 230 20 of of ADP bracis-28396 230 21 wood wood NOUN bracis-28396 230 22 boards board NOUN bracis-28396 230 23 in in ADP bracis-28396 230 24 the the DET bracis-28396 230 25 sawmill sawmill NOUN bracis-28396 230 26 industry industry NOUN bracis-28396 230 27 . . PUNCT bracis-28396 231 1 experiments experiment NOUN bracis-28396 231 2 showed show VERB bracis-28396 231 3 that that SCONJ bracis-28396 231 4 a a DET bracis-28396 231 5 solution solution NOUN bracis-28396 231 6 based base VERB bracis-28396 231 7 on on ADP bracis-28396 231 8 pso pso NOUN bracis-28396 231 9 led lead VERB bracis-28396 231 10 to to ADP bracis-28396 231 11 satisfactory satisfactory ADJ bracis-28396 231 12 results result NOUN bracis-28396 231 13 compared compare VERB bracis-28396 231 14 to to ADP bracis-28396 231 15 baseline baseline VERB bracis-28396 231 16 methods method NOUN bracis-28396 231 17 . . PUNCT bracis-28396 232 1 according accord VERB bracis-28396 232 2 to to ADP bracis-28396 232 3 a a DET bracis-28396 232 4 statistical statistical ADJ bracis-28396 232 5 test test NOUN bracis-28396 232 6 , , PUNCT bracis-28396 232 7 results result NOUN bracis-28396 232 8 show show VERB bracis-28396 232 9 a a DET bracis-28396 232 10 significant significant ADJ bracis-28396 232 11 difference difference NOUN bracis-28396 232 12 during during ADP bracis-28396 232 13 the the DET bracis-28396 232 14 optimization optimization NOUN bracis-28396 232 15 task task NOUN bracis-28396 232 16 but but CCONJ bracis-28396 232 17 not not PART bracis-28396 232 18 for for ADP bracis-28396 232 19 the the DET bracis-28396 232 20 generalization generalization NOUN bracis-28396 232 21 phase phase NOUN bracis-28396 232 22 . . PUNCT bracis-28396 233 1 these these DET bracis-28396 233 2 experimental experimental ADJ bracis-28396 233 3 results result NOUN bracis-28396 233 4 suggest suggest VERB bracis-28396 233 5 that that SCONJ bracis-28396 233 6 mlp mlp PROPN bracis-28396 233 7 hyperparameter hyperparameter NOUN bracis-28396 233 8 tuning tuning NOUN bracis-28396 233 9 and and CCONJ bracis-28396 233 10 feature feature NOUN bracis-28396 233 11 selection selection NOUN bracis-28396 233 12 are be AUX bracis-28396 233 13 essential essential ADJ bracis-28396 233 14 to to PART bracis-28396 233 15 obtain obtain VERB bracis-28396 233 16 models model NOUN bracis-28396 233 17 with with ADP bracis-28396 233 18 higher high ADJ bracis-28396 233 19 predictive predictive ADJ bracis-28396 233 20 performance performance NOUN bracis-28396 233 21 . . PUNCT bracis-28396 234 1 also also ADV bracis-28396 234 2 , , PUNCT bracis-28396 234 3 one one PRON bracis-28396 234 4 can can AUX bracis-28396 234 5 notice notice VERB bracis-28396 234 6 that that SCONJ bracis-28396 234 7 these these DET bracis-28396 234 8 tasks task NOUN bracis-28396 234 9 are be AUX bracis-28396 234 10 interdependent interdependent ADJ bracis-28396 234 11 since since SCONJ bracis-28396 234 12 the the DET bracis-28396 234 13 hyperparameter hyperparameter NOUN bracis-28396 234 14 values value NOUN bracis-28396 234 15 should should AUX bracis-28396 234 16 be be AUX bracis-28396 234 17 adjusted adjust VERB bracis-28396 234 18 according accord VERB bracis-28396 234 19 to to ADP bracis-28396 234 20 a a DET bracis-28396 234 21 subset subset NOUN bracis-28396 234 22 of of ADP bracis-28396 234 23 features feature NOUN bracis-28396 234 24 and and CCONJ bracis-28396 234 25 vice vice ADV bracis-28396 234 26 versa versa ADV bracis-28396 234 27 . . PUNCT bracis-28396 235 1 consequently consequently ADV bracis-28396 235 2 , , PUNCT bracis-28396 235 3 for for ADP bracis-28396 235 4 the the DET bracis-28396 235 5 problem problem NOUN bracis-28396 235 6 investigated investigate VERB bracis-28396 235 7 in in ADP bracis-28396 235 8 this this DET bracis-28396 235 9 work work NOUN bracis-28396 235 10 , , PUNCT bracis-28396 235 11 performing perform VERB bracis-28396 235 12 only only ADV bracis-28396 235 13 one one NUM bracis-28396 235 14 of of ADP bracis-28396 235 15 them they PRON bracis-28396 235 16 was be AUX bracis-28396 235 17 enough enough ADJ bracis-28396 235 18 to to PART bracis-28396 235 19 reach reach VERB bracis-28396 235 20 substantial substantial ADJ bracis-28396 235 21 gain gain NOUN bracis-28396 235 22 . . PUNCT bracis-28396 236 1 finally finally ADV bracis-28396 236 2 , , PUNCT bracis-28396 236 3 the the DET bracis-28396 236 4 accuracy accuracy NOUN bracis-28396 236 5 obtained obtain VERB bracis-28396 236 6 in in ADP bracis-28396 236 7 this this DET bracis-28396 236 8 study study NOUN bracis-28396 236 9 supports support VERB bracis-28396 236 10 employing employ VERB bracis-28396 236 11 machine machine NOUN bracis-28396 236 12 learning learning NOUN bracis-28396 236 13 models model NOUN bracis-28396 236 14 for for ADP bracis-28396 236 15 industrial industrial ADJ bracis-28396 236 16 implementation implementation NOUN bracis-28396 236 17 , , PUNCT bracis-28396 236 18 contributing contribute VERB bracis-28396 236 19 to to ADP bracis-28396 236 20 overall overall ADJ bracis-28396 236 21 cost cost NOUN bracis-28396 236 22 reduction reduction NOUN bracis-28396 236 23 and and CCONJ bracis-28396 236 24 improvement improvement NOUN bracis-28396 236 25 in in ADP bracis-28396 236 26 competitiveness competitiveness NOUN bracis-28396 236 27 . . PUNCT bracis-28396 237 1 regarding regard VERB bracis-28396 237 2 future future ADJ bracis-28396 237 3 works work NOUN bracis-28396 237 4 , , PUNCT bracis-28396 237 5 we we PRON bracis-28396 237 6 intend intend VERB bracis-28396 237 7 to to PART bracis-28396 237 8 perform perform VERB bracis-28396 237 9 a a DET bracis-28396 237 10 transfer transfer NOUN bracis-28396 237 11 learning learn VERB bracis-28396 237 12 from from ADP bracis-28396 237 13 a a DET bracis-28396 237 14 cnn cnn NOUN bracis-28396 237 15 trained train VERB bracis-28396 237 16 using use VERB bracis-28396 237 17 a a DET bracis-28396 237 18 dataset dataset NOUN bracis-28396 237 19 composed compose VERB bracis-28396 237 20 of of ADP bracis-28396 237 21 a a DET bracis-28396 237 22 more more ADV bracis-28396 237 23 substantial substantial ADJ bracis-28396 237 24 number number NOUN bracis-28396 237 25 of of ADP bracis-28396 237 26 wood wood NOUN bracis-28396 237 27 image image NOUN bracis-28396 237 28 samples sample NOUN bracis-28396 237 29 . . PUNCT bracis-28396 238 1 besides besides SCONJ bracis-28396 238 2 , , PUNCT bracis-28396 238 3 we we PRON bracis-28396 238 4 are be AUX bracis-28396 238 5 willing willing ADJ bracis-28396 238 6 to to PART bracis-28396 238 7 investigate investigate VERB bracis-28396 238 8 and and CCONJ bracis-28396 238 9 compare compare VERB bracis-28396 238 10 different different ADJ bracis-28396 238 11 image image NOUN bracis-28396 238 12 descriptors descriptor NOUN bracis-28396 238 13 , , PUNCT bracis-28396 238 14 non non ADJ bracis-28396 238 15 - - ADJ bracis-28396 238 16 linear linear ADJ bracis-28396 238 17 data datum NOUN bracis-28396 238 18 reduction reduction NOUN bracis-28396 238 19 techniques technique NOUN bracis-28396 238 20 , , PUNCT bracis-28396 238 21 and and CCONJ bracis-28396 238 22 deep deep ADJ bracis-28396 238 23 learning learning NOUN bracis-28396 238 24 models model NOUN bracis-28396 238 25 . . PUNCT bracis-28396 239 1 notes note VERB bracis-28396 239 2 1.https://www.python.org/. 1.https://www.python.org/. PROPN bracis-28396 239 3 2.http://scikit-image.org/. 2.http://scikit-image.org/. NUM bracis-28396 239 4 references reference NOUN bracis-28396 239 5 abdullah abdullah PROPN bracis-28396 239 6 , , PUNCT bracis-28396 239 7 a. a. PROPN bracis-28396 239 8 , , PUNCT bracis-28396 239 9 ismail ismail PROPN bracis-28396 239 10 , , PUNCT bracis-28396 239 11 n.k.n n.k.n PROPN bracis-28396 239 12 . . PROPN bracis-28396 239 13 , , PUNCT bracis-28396 239 14 kadir kadir PROPN bracis-28396 239 15 , , PUNCT bracis-28396 239 16 t.a.a t.a.a PROPN bracis-28396 239 17 . . PROPN bracis-28396 239 18 , , PUNCT bracis-28396 239 19 zain zain PROPN bracis-28396 239 20 , , PUNCT bracis-28396 239 21 j.m j.m PROPN bracis-28396 239 22 . . PROPN bracis-28396 239 23 , , PUNCT bracis-28396 239 24 jusoh jusoh PROPN bracis-28396 239 25 , , PUNCT bracis-28396 239 26 n.a n.a PROPN bracis-28396 239 27 . . PROPN bracis-28396 239 28 , , PUNCT bracis-28396 239 29 ali ali PROPN bracis-28396 239 30 , , PUNCT bracis-28396 239 31 n.m n.m PROPN bracis-28396 239 32 . . PROPN bracis-28396 239 33 : : PUNCT bracis-28396 239 34 agar agar NOUN bracis-28396 239 35 wood wood NOUN bracis-28396 239 36 grade grade NOUN bracis-28396 239 37 determination determination NOUN bracis-28396 239 38 system system NOUN bracis-28396 239 39 using use VERB bracis-28396 239 40 image image NOUN bracis-28396 239 41 processing processing NOUN bracis-28396 239 42 technique technique NOUN bracis-28396 239 43 . . PUNCT bracis-28396 240 1 in in ADP bracis-28396 240 2 : : PUNCT bracis-28396 240 3 proceedings proceeding NOUN bracis-28396 240 4 of of ADP bracis-28396 240 5 the the DET bracis-28396 240 6 international international ADJ bracis-28396 240 7 conference conference NOUN bracis-28396 240 8 on on ADP bracis-28396 240 9 electrical electrical ADJ bracis-28396 240 10 engineering engineering NOUN bracis-28396 240 11 and and CCONJ bracis-28396 240 12 informatics informatics PROPN bracis-28396 240 13 institut institut PROPN bracis-28396 240 14 teknologi teknologi PROPN bracis-28396 240 15 bandung bandung PROPN bracis-28396 240 16 ( ( PUNCT bracis-28396 240 17 2007 2007 NUM bracis-28396 240 18 ) ) PUNCT bracis-28396 240 19 google google PROPN bracis-28396 240 20 scholar scholar NOUN bracis-28396 240 21     SPACE bracis-28396 240 22 affonso affonso PROPN bracis-28396 240 23 , , PUNCT bracis-28396 240 24 c. c. PROPN bracis-28396 240 25 , , PUNCT bracis-28396 240 26 rossi rossi ADJ bracis-28396 240 27 , , PUNCT bracis-28396 240 28 a.l.d a.l.d NOUN bracis-28396 240 29 . . PROPN bracis-28396 240 30 , , PUNCT bracis-28396 240 31 vieira vieira PROPN bracis-28396 240 32 , , PUNCT bracis-28396 240 33 f.h.a f.h.a PROPN bracis-28396 240 34 . . PROPN bracis-28396 240 35 , , PUNCT bracis-28396 240 36 carvalho carvalho PROPN bracis-28396 240 37 , , PUNCT bracis-28396 240 38 a.c.p.l.f a.c.p.l.f PROPN bracis-28396 240 39 . . PUNCT bracis-28396 240 40 : : PUNCT bracis-28396 241 1 deep deep ADJ bracis-28396 241 2 learning learning NOUN bracis-28396 241 3 for for ADP bracis-28396 241 4 biological biological ADJ bracis-28396 241 5 image image NOUN bracis-28396 241 6 classification classification NOUN bracis-28396 241 7 . . PUNCT bracis-28396 242 1 expert expert NOUN bracis-28396 242 2 syst syst PROPN bracis-28396 242 3 . . PUNCT bracis-28396 243 1 appl appl PROPN bracis-28396 243 2 . . PROPN bracis-28396 243 3 85 85 NUM bracis-28396 243 4 , , PUNCT bracis-28396 243 5 114–122 114–122 NUM bracis-28396 243 6 ( ( PUNCT bracis-28396 243 7 2017 2017 NUM bracis-28396 243 8 ) ) PUNCT bracis-28396 243 9 . . PUNCT bracis-28396 244 1 https://doi.org/10.1016/j.eswa.2017.05.039 https://doi.org/10.1016/j.eswa.2017.05.039 NUM bracis-28396 244 2 brodersen brodersen NOUN bracis-28396 244 3 , , PUNCT bracis-28396 244 4 k.h k.h PROPN bracis-28396 244 5 . . PROPN bracis-28396 244 6 , , PUNCT bracis-28396 244 7 ong ong PROPN bracis-28396 244 8 , , PUNCT bracis-28396 244 9 c.s c.s PROPN bracis-28396 244 10 . . PROPN bracis-28396 244 11 , , PUNCT bracis-28396 244 12 stephan stephan PROPN bracis-28396 244 13 , , PUNCT bracis-28396 244 14 k.e k.e PROPN bracis-28396 244 15 . . PROPN bracis-28396 244 16 , , PUNCT bracis-28396 244 17 buhmann buhmann PROPN bracis-28396 244 18 , , PUNCT bracis-28396 244 19 j.m j.m PROPN bracis-28396 244 20 . . PROPN bracis-28396 244 21 : : PUNCT bracis-28396 245 1 the the DET bracis-28396 245 2 balanced balanced ADJ bracis-28396 245 3 accuracy accuracy NOUN bracis-28396 245 4 and and CCONJ bracis-28396 245 5 its its PRON bracis-28396 245 6 posterior posterior ADJ bracis-28396 245 7 distribution distribution NOUN bracis-28396 245 8 . . PUNCT bracis-28396 246 1 in in ADP bracis-28396 246 2 : : PUNCT bracis-28396 246 3 2010 2010 NUM bracis-28396 246 4 20th 20th ADJ bracis-28396 246 5 international international ADJ bracis-28396 246 6 conference conference NOUN bracis-28396 246 7 on on ADP bracis-28396 246 8 pattern pattern NOUN bracis-28396 246 9 recognition recognition NOUN bracis-28396 246 10 , , PUNCT bracis-28396 246 11 pp pp PROPN bracis-28396 246 12 . . PUNCT bracis-28396 246 13 3121–3124 3121–3124 NUM bracis-28396 246 14 , , PUNCT bracis-28396 246 15 august august PROPN bracis-28396 246 16 2010 2010 NUM bracis-28396 246 17 . . PUNCT bracis-28396 247 1 https://doi.org/10.1109/icpr.2010.764 https://doi.org/10.1109/icpr.2010.764 PROPN bracis-28396 247 2 cao cao PROPN bracis-28396 247 3 , , PUNCT bracis-28396 247 4 y. y. PROPN bracis-28396 247 5 , , PUNCT bracis-28396 247 6 et et PROPN bracis-28396 247 7 al al PROPN bracis-28396 247 8 . . PUNCT bracis-28396 247 9 : : PUNCT bracis-28396 248 1 a a DET bracis-28396 248 2 new new ADJ bracis-28396 248 3 intelligence intelligence NOUN bracis-28396 248 4 fuzzy fuzzy ADV bracis-28396 248 5 - - PUNCT bracis-28396 248 6 based base VERB bracis-28396 248 7 hybrid hybrid ADJ bracis-28396 248 8 metaheuristic metaheuristic ADJ bracis-28396 248 9 algorithm algorithm NOUN bracis-28396 248 10 for for ADP bracis-28396 248 11 analyzing analyze VERB bracis-28396 248 12 the the DET bracis-28396 248 13 application application NOUN bracis-28396 248 14 of of ADP bracis-28396 248 15 tea tea NOUN bracis-28396 248 16 waste waste NOUN bracis-28396 248 17 in in ADP bracis-28396 248 18 concrete concrete NOUN bracis-28396 248 19 as as ADP bracis-28396 248 20 natural natural ADJ bracis-28396 248 21 fiber fiber NOUN bracis-28396 248 22 . . PUNCT bracis-28396 249 1 comput comput NOUN bracis-28396 249 2 . . PUNCT bracis-28396 250 1 electron electron PROPN bracis-28396 250 2 . . PUNCT bracis-28396 251 1 agric agric PROPN bracis-28396 251 2 . . PROPN bracis-28396 252 1 190 190 NUM bracis-28396 252 2 , , PUNCT bracis-28396 252 3 106420 106420 NUM bracis-28396 252 4 ( ( PUNCT bracis-28396 252 5 2021 2021 NUM bracis-28396 252 6 ) ) PUNCT bracis-28396 252 7 article article NOUN bracis-28396 252 8     SPACE bracis-28396 252 9 google google PROPN bracis-28396 252 10 scholar scholar NOUN bracis-28396 252 11     SPACE bracis-28396 252 12 de de X bracis-28396 252 13 souza souza PROPN bracis-28396 252 14 , , PUNCT bracis-28396 252 15 l.a l.a PROPN bracis-28396 252 16 . . PROPN bracis-28396 252 17 , , PUNCT bracis-28396 252 18 et et PROPN bracis-28396 252 19 al al PROPN bracis-28396 252 20 . . PUNCT bracis-28396 252 21 : : PUNCT bracis-28396 253 1 fine fine ADJ bracis-28396 253 2 - - PUNCT bracis-28396 253 3 tuning tune VERB bracis-28396 253 4 generative generative ADJ bracis-28396 253 5 adversarial adversarial ADJ bracis-28396 253 6 networks network NOUN bracis-28396 253 7 using use VERB bracis-28396 253 8 metaheuristics metaheuristic NOUN bracis-28396 253 9 - - PUNCT bracis-28396 253 10 a a DET bracis-28396 253 11 case case NOUN bracis-28396 253 12 study study NOUN bracis-28396 253 13 on on ADP bracis-28396 253 14 barrett barrett PROPN bracis-28396 253 15 ’s ’s PART bracis-28396 253 16 esophagus esophagus PROPN bracis-28396 253 17 identification identification NOUN bracis-28396 253 18 . . PUNCT bracis-28396 254 1 in in ADP bracis-28396 254 2 : : PUNCT bracis-28396 254 3 bildverarbeitung bildverarbeitung PROPN bracis-28396 254 4 für für PROPN bracis-28396 254 5 die die VERB bracis-28396 254 6 medizin medizin PROPN bracis-28396 254 7 , , PUNCT bracis-28396 254 8 pp pp ADP bracis-28396 254 9 . . PUNCT bracis-28396 255 1 205–210 205–210 NUM bracis-28396 255 2 ( ( PUNCT bracis-28396 255 3 2021 2021 NUM bracis-28396 255 4 ) ) PUNCT bracis-28396 255 5 google google NOUN bracis-28396 255 6 scholar scholar NOUN bracis-28396 255 7     SPACE bracis-28396 255 8 gu gu NOUN bracis-28396 255 9 , , PUNCT bracis-28396 255 10 i.y.h i.y.h PROPN bracis-28396 255 11 . . PROPN bracis-28396 255 12 , , PUNCT bracis-28396 255 13 andersson andersson PROPN bracis-28396 255 14 , , PUNCT bracis-28396 255 15 h. h. PROPN bracis-28396 255 16 , , PUNCT bracis-28396 255 17 vicen vicen PROPN bracis-28396 255 18 , , PUNCT bracis-28396 255 19 r. r. PROPN bracis-28396 255 20 : : PUNCT bracis-28396 255 21 automatic automatic ADJ bracis-28396 255 22 classification classification NOUN bracis-28396 255 23 of of ADP bracis-28396 255 24 wood wood NOUN bracis-28396 255 25 defects defect NOUN bracis-28396 255 26 using use VERB bracis-28396 255 27 support support NOUN bracis-28396 255 28 vector vector NOUN bracis-28396 255 29 machines machine NOUN bracis-28396 255 30 . . PUNCT bracis-28396 256 1 in in ADP bracis-28396 256 2 : : PUNCT bracis-28396 256 3 bolc bolc PROPN bracis-28396 256 4 , , PUNCT bracis-28396 256 5 l. l. PROPN bracis-28396 256 6 , , PUNCT bracis-28396 256 7 kulikowski kulikowski PROPN bracis-28396 256 8 , , PUNCT bracis-28396 256 9 j.l j.l PROPN bracis-28396 256 10 . . PROPN bracis-28396 256 11 , , PUNCT bracis-28396 256 12 wojciechowski wojciechowski PROPN bracis-28396 256 13 , , PUNCT bracis-28396 256 14 k. k. PROPN bracis-28396 256 15 ( ( PUNCT bracis-28396 256 16 eds eds PROPN bracis-28396 256 17 . . PUNCT bracis-28396 256 18 ) ) PUNCT bracis-28396 256 19 iccvg iccvg ADJ bracis-28396 256 20 2008 2008 NUM bracis-28396 256 21 . . PUNCT bracis-28396 257 1 lncs lncs PROPN bracis-28396 257 2 , , PUNCT bracis-28396 257 3 vol vol NOUN bracis-28396 257 4 . . PROPN bracis-28396 257 5 5337 5337 NUM bracis-28396 257 6 , , PUNCT bracis-28396 257 7 pp pp ADJ bracis-28396 257 8 . . PUNCT bracis-28396 258 1 356–367 356–367 NUM bracis-28396 258 2 . . PUNCT bracis-28396 258 3 springer springer NOUN bracis-28396 258 4 , , PUNCT bracis-28396 258 5 heidelberg heidelberg PROPN bracis-28396 258 6 ( ( PUNCT bracis-28396 258 7 2009 2009 NUM bracis-28396 258 8 ) ) PUNCT bracis-28396 258 9 . . PUNCT bracis-28396 259 1 https://doi.org/10.1007/978-3-642-02345-3_35 https://doi.org/10.1007/978-3-642-02345-3_35 PROPN bracis-28396 259 2 chapter chapter NOUN bracis-28396 259 3     SPACE bracis-28396 259 4 google google PROPN bracis-28396 259 5 scholar scholar NOUN bracis-28396 259 6     SPACE bracis-28396 259 7 hall hall NOUN bracis-28396 259 8 , , PUNCT bracis-28396 259 9 m. m. NOUN bracis-28396 259 10 , , PUNCT bracis-28396 259 11 frank frank PROPN bracis-28396 259 12 , , PUNCT bracis-28396 259 13 e. e. PROPN bracis-28396 259 14 , , PUNCT bracis-28396 259 15 holmes holmes PROPN bracis-28396 259 16 , , PUNCT bracis-28396 259 17 g. g. PROPN bracis-28396 259 18 , , PUNCT bracis-28396 259 19 pfahringer pfahringer NOUN bracis-28396 259 20 , , PUNCT bracis-28396 259 21 b. b. PROPN bracis-28396 259 22 , , PUNCT bracis-28396 259 23 reutemann reutemann PROPN bracis-28396 259 24 , , PUNCT bracis-28396 259 25 p. p. PROPN bracis-28396 259 26 , , PUNCT bracis-28396 259 27 witten witten PROPN bracis-28396 259 28 , , PUNCT bracis-28396 259 29 i.h i.h PROPN bracis-28396 259 30 . . PROPN bracis-28396 259 31 : : PUNCT bracis-28396 260 1 the the DET bracis-28396 260 2 weka weka PROPN bracis-28396 260 3 data data PROPN bracis-28396 260 4 mining mining NOUN bracis-28396 260 5 software software NOUN bracis-28396 260 6 : : PUNCT bracis-28396 260 7 an an DET bracis-28396 260 8 update update NOUN bracis-28396 260 9 . . PUNCT bracis-28396 261 1 sigkdd sigkdd ADJ bracis-28396 261 2 explor explor PROPN bracis-28396 261 3 . . PUNCT bracis-28396 262 1 newsl newsl PROPN bracis-28396 262 2 . . PUNCT bracis-28396 263 1 11(1 11(1 NUM bracis-28396 263 2 ) ) PUNCT bracis-28396 263 3 , , PUNCT bracis-28396 263 4 10–18 10–18 NUM bracis-28396 263 5 ( ( PUNCT bracis-28396 263 6 2009 2009 NUM bracis-28396 263 7 ) ) PUNCT bracis-28396 263 8 . . PUNCT bracis-28396 264 1 https://doi.org/10.1145/1656274.1656278 https://doi.org/10.1145/1656274.1656278 NOUN bracis-28396 264 2 article article NOUN bracis-28396 264 3     SPACE bracis-28396 264 4 google google PROPN bracis-28396 264 5 scholar scholar NOUN bracis-28396 264 6     SPACE bracis-28396 264 7 haralick haralick NOUN bracis-28396 264 8 , , PUNCT bracis-28396 264 9 r. r. PROPN bracis-28396 264 10 , , PUNCT bracis-28396 264 11 shanmugam shanmugam PROPN bracis-28396 264 12 , , PUNCT bracis-28396 264 13 k. k. PROPN bracis-28396 264 14 , , PUNCT bracis-28396 264 15 distein distein PROPN bracis-28396 264 16 , , PUNCT bracis-28396 264 17 i. i. NOUN bracis-28396 264 18 : : PUNCT bracis-28396 264 19 textual textual ADJ bracis-28396 264 20 features feature VERB bracis-28396 264 21 for for ADP bracis-28396 264 22 image image NOUN bracis-28396 264 23 classification classification NOUN bracis-28396 264 24 . . PUNCT bracis-28396 265 1 ieee ieee PROPN bracis-28396 265 2 trans trans PROPN bracis-28396 265 3 . . PUNCT bracis-28396 266 1 syst syst PROPN bracis-28396 266 2 . . PUNCT bracis-28396 267 1 man man PROPN bracis-28396 267 2 cybern cybern PROPN bracis-28396 267 3 . . PUNCT bracis-28396 268 1 smc smc PROPN bracis-28396 268 2 3(6 3(6 PROPN bracis-28396 268 3 ) ) PUNCT bracis-28396 268 4 , , PUNCT bracis-28396 268 5 610–621 610–621 NUM bracis-28396 268 6 ( ( PUNCT bracis-28396 268 7 1973 1973 NUM bracis-28396 268 8 ) ) PUNCT bracis-28396 268 9 google google PROPN bracis-28396 268 10 scholar scholar NOUN bracis-28396 268 11     SPACE bracis-28396 268 12 kennedy kennedy PROPN bracis-28396 268 13 , , PUNCT bracis-28396 268 14 j. j. PROPN bracis-28396 268 15 , , PUNCT bracis-28396 268 16 eberhart eberhart PROPN bracis-28396 268 17 , , PUNCT bracis-28396 268 18 r.c r.c PROPN bracis-28396 268 19 . . PROPN bracis-28396 268 20 : : PUNCT bracis-28396 269 1 a a DET bracis-28396 269 2 discrete discrete ADJ bracis-28396 269 3 binary binary ADJ bracis-28396 269 4 version version NOUN bracis-28396 269 5 of of ADP bracis-28396 269 6 the the DET bracis-28396 269 7 particle particle NOUN bracis-28396 269 8 swarm swarm NOUN bracis-28396 269 9 algorithm algorithm NOUN bracis-28396 269 10 . . PUNCT bracis-28396 270 1 in in ADP bracis-28396 270 2 : : PUNCT bracis-28396 270 3 1997 1997 NUM bracis-28396 270 4 ieee ieee PROPN bracis-28396 270 5 international international ADJ bracis-28396 270 6 conference conference NOUN bracis-28396 270 7 on on ADP bracis-28396 270 8 systems system NOUN bracis-28396 270 9 , , PUNCT bracis-28396 270 10 man man NOUN bracis-28396 270 11 , , PUNCT bracis-28396 270 12 and and CCONJ bracis-28396 270 13 cybernetics cybernetic NOUN bracis-28396 270 14 . . PUNCT bracis-28396 271 1 computational computational ADJ bracis-28396 271 2 cybernetics cybernetic NOUN bracis-28396 271 3 and and CCONJ bracis-28396 271 4 simulation simulation NOUN bracis-28396 271 5 , , PUNCT bracis-28396 271 6 vol vol NOUN bracis-28396 271 7 . . PROPN bracis-28396 271 8 5 5 NUM bracis-28396 271 9 , , PUNCT bracis-28396 271 10 pp pp ADJ bracis-28396 271 11 . . PUNCT bracis-28396 271 12 4104–4108 4104–4108 NUM bracis-28396 271 13 , , PUNCT bracis-28396 271 14 october october PROPN bracis-28396 271 15 1997 1997 NUM bracis-28396 271 16 . . PUNCT bracis-28396 272 1 https://doi.org/10.1109/icsmc.1997.637339 https://doi.org/10.1109/icsmc.1997.637339 PROPN bracis-28396 272 2 kennedy kennedy PROPN bracis-28396 272 3 , , PUNCT bracis-28396 272 4 j. j. PROPN bracis-28396 272 5 , , PUNCT bracis-28396 272 6 eberhart eberhart PROPN bracis-28396 272 7 , , PUNCT bracis-28396 272 8 r. r. PROPN bracis-28396 272 9 : : PUNCT bracis-28396 272 10 swarm swarm NOUN bracis-28396 272 11 intelligence intelligence NOUN bracis-28396 272 12 . . PUNCT bracis-28396 273 1 morgan morgan PROPN bracis-28396 273 2 kaufmann kaufmann PROPN bracis-28396 273 3 publishers publishers PROPN bracis-28396 273 4 ( ( PUNCT bracis-28396 273 5 2001 2001 NUM bracis-28396 273 6 ) ) PUNCT bracis-28396 273 7 google google PROPN bracis-28396 273 8 scholar scholar NOUN bracis-28396 273 9     SPACE bracis-28396 273 10 kennedy kennedy PROPN bracis-28396 273 11 , , PUNCT bracis-28396 273 12 j. j. PROPN bracis-28396 273 13 , , PUNCT bracis-28396 273 14 eberhart eberhart PROPN bracis-28396 273 15 , , PUNCT bracis-28396 273 16 r. r. PROPN bracis-28396 273 17 : : PUNCT bracis-28396 273 18 particle particle NOUN bracis-28396 273 19 swarm swarm NOUN bracis-28396 273 20 optimization optimization NOUN bracis-28396 273 21 . . PUNCT bracis-28396 274 1 in in ADP bracis-28396 274 2 : : PUNCT bracis-28396 274 3 proceedings proceeding NOUN bracis-28396 274 4 of of ADP bracis-28396 274 5 the the DET bracis-28396 274 6 ieee ieee NOUN bracis-28396 274 7 international international PROPN bracis-28396 274 8 conference conference NOUN bracis-28396 274 9 on on ADP bracis-28396 274 10 neural neural ADJ bracis-28396 274 11 networks network NOUN bracis-28396 274 12 , , PUNCT bracis-28396 274 13 vol vol NOUN bracis-28396 274 14 . . PROPN bracis-28396 274 15 4 4 NUM bracis-28396 274 16 , , PUNCT bracis-28396 274 17 pp pp ADJ bracis-28396 274 18 . . PUNCT bracis-28396 274 19 1942–1948 1942–1948 NUM bracis-28396 274 20 . . PUNCT bracis-28396 275 1 perth perth PROPN bracis-28396 275 2 , , PUNCT bracis-28396 275 3 australia australia PROPN bracis-28396 275 4 ( ( PUNCT bracis-28396 275 5 1995 1995 NUM bracis-28396 275 6 ) ) PUNCT bracis-28396 275 7 google google PROPN bracis-28396 275 8 scholar scholar NOUN bracis-28396 275 9     SPACE bracis-28396 275 10 luo luo PROPN bracis-28396 275 11 , , PUNCT bracis-28396 275 12 g. g. PROPN bracis-28396 275 13 : : PUNCT bracis-28396 275 14 a a DET bracis-28396 275 15 review review NOUN bracis-28396 275 16 of of ADP bracis-28396 275 17 automatic automatic ADJ bracis-28396 275 18 selection selection NOUN bracis-28396 275 19 methods method NOUN bracis-28396 275 20 for for ADP bracis-28396 275 21 machine machine NOUN bracis-28396 275 22 learning learn VERB bracis-28396 275 23 algorithms algorithm NOUN bracis-28396 275 24 and and CCONJ bracis-28396 275 25 hyper hyper ADJ bracis-28396 275 26 - - ADJ bracis-28396 275 27 parameter parameter ADJ bracis-28396 275 28 values value NOUN bracis-28396 275 29 . . PUNCT bracis-28396 276 1 netw netw NOUN bracis-28396 276 2 . . PUNCT bracis-28396 277 1 model model PROPN bracis-28396 277 2 . . PUNCT bracis-28396 278 1 anal anal PROPN bracis-28396 278 2 . . PUNCT bracis-28396 279 1 health health NOUN bracis-28396 279 2 inform inform NOUN bracis-28396 279 3 . . PUNCT bracis-28396 280 1 bioinforma bioinforma NOUN bracis-28396 280 2 . . PUNCT bracis-28396 281 1 5(1 5(1 NUM bracis-28396 281 2 ) ) PUNCT bracis-28396 281 3 , , PUNCT bracis-28396 281 4 1–16 1–16 NOUN bracis-28396 281 5 ( ( PUNCT bracis-28396 281 6 2016 2016 NUM bracis-28396 281 7 ) ) PUNCT bracis-28396 281 8 . . PUNCT bracis-28396 282 1 https://doi.org/10.1007/s13721-016-0125-6 https://doi.org/10.1007/s13721-016-0125-6 PROPN bracis-28396 282 2 article article NOUN bracis-28396 282 3     SPACE bracis-28396 282 4 google google PROPN bracis-28396 282 5 scholar scholar NOUN bracis-28396 282 6     SPACE bracis-28396 282 7 ojala ojala NOUN bracis-28396 282 8 , , PUNCT bracis-28396 282 9 t. t. PROPN bracis-28396 282 10 , , PUNCT bracis-28396 282 11 pietikainen pietikainen NOUN bracis-28396 282 12 , , PUNCT bracis-28396 282 13 m. m. NOUN bracis-28396 282 14 , , PUNCT bracis-28396 282 15 harwood harwood NOUN bracis-28396 282 16 , , PUNCT bracis-28396 282 17 d. d. PROPN bracis-28396 282 18 : : PUNCT bracis-28396 282 19 comparative comparative ADJ bracis-28396 282 20 study study NOUN bracis-28396 282 21 of of ADP bracis-28396 282 22 texture texture ADJ bracis-28396 282 23 measures measure NOUN bracis-28396 282 24 with with ADP bracis-28396 282 25 classification classification NOUN bracis-28396 282 26 based base VERB bracis-28396 282 27 on on ADP bracis-28396 282 28 feature feature NOUN bracis-28396 282 29 distributions distribution NOUN bracis-28396 282 30 . . PUNCT bracis-28396 283 1 pattern pattern NOUN bracis-28396 283 2 recogn recogn VERB bracis-28396 283 3 . . PROPN bracis-28396 283 4 , , PUNCT bracis-28396 283 5 51–59 51–59 NUM bracis-28396 283 6 ( ( PUNCT bracis-28396 283 7 1996 1996 NUM bracis-28396 283 8 ) ) PUNCT bracis-28396 283 9 google google PROPN bracis-28396 283 10 scholar scholar NOUN bracis-28396 283 11     SPACE bracis-28396 283 12 passos passos PROPN bracis-28396 283 13 , , PUNCT bracis-28396 283 14 l.a l.a PROPN bracis-28396 283 15 . . PROPN bracis-28396 283 16 , , PUNCT bracis-28396 283 17 paulo paulo PROPN bracis-28396 283 18 papa papa PROPN bracis-28396 283 19 , , PUNCT bracis-28396 283 20 j. j. PROPN bracis-28396 283 21 : : PUNCT bracis-28396 283 22 fine fine ADJ bracis-28396 283 23 - - PUNCT bracis-28396 283 24 tuning tuning NOUN bracis-28396 283 25 infinity infinity NOUN bracis-28396 283 26 restricted restrict VERB bracis-28396 283 27 boltzmann boltzmann PROPN bracis-28396 283 28 machines machine NOUN bracis-28396 283 29 . . PUNCT bracis-28396 284 1 in in ADP bracis-28396 284 2 : : PUNCT bracis-28396 284 3 2017 2017 NUM bracis-28396 284 4 30th 30th NOUN bracis-28396 284 5 sibgrapi sibgrapi ADJ bracis-28396 284 6 conference conference NOUN bracis-28396 284 7 on on ADP bracis-28396 284 8 graphics graphic NOUN bracis-28396 284 9 , , PUNCT bracis-28396 284 10 patterns pattern NOUN bracis-28396 284 11 and and CCONJ bracis-28396 284 12 images image NOUN bracis-28396 284 13 ( ( PUNCT bracis-28396 284 14 sibgrapi sibgrapi ADJ bracis-28396 284 15 ) ) PUNCT bracis-28396 284 16 , , PUNCT bracis-28396 284 17 pp pp PROPN bracis-28396 284 18 . . PUNCT bracis-28396 284 19 63–70 63–70 NOUN bracis-28396 284 20 . . PUNCT bracis-28396 285 1 ieee ieee NOUN bracis-28396 285 2 ( ( PUNCT bracis-28396 285 3 2017 2017 NUM bracis-28396 285 4 ) ) PUNCT bracis-28396 285 5 google google PROPN bracis-28396 285 6 scholar scholar NOUN bracis-28396 285 7     SPACE bracis-28396 285 8 pereira pereira PROPN bracis-28396 285 9 , , PUNCT bracis-28396 285 10 c.r c.r PROPN bracis-28396 285 11 . . PROPN bracis-28396 285 12 , , PUNCT bracis-28396 285 13 passos passos PROPN bracis-28396 285 14 , , PUNCT bracis-28396 285 15 l.a l.a PROPN bracis-28396 285 16 . . PROPN bracis-28396 285 17 , , PUNCT bracis-28396 285 18 rodrigues rodrigues PROPN bracis-28396 285 19 , , PUNCT bracis-28396 285 20 d. d. PROPN bracis-28396 285 21 , , PUNCT bracis-28396 285 22 de de PROPN bracis-28396 285 23 souza souza PROPN bracis-28396 285 24 , , PUNCT bracis-28396 285 25 a.n a.n PROPN bracis-28396 285 26 . . PROPN bracis-28396 285 27 , , PUNCT bracis-28396 285 28 papa papa PROPN bracis-28396 285 29 , , PUNCT bracis-28396 285 30 j.p j.p PROPN bracis-28396 285 31 . . PROPN bracis-28396 285 32 : : PUNCT bracis-28396 285 33 jade jade NOUN bracis-28396 285 34 - - PUNCT bracis-28396 285 35 based base VERB bracis-28396 285 36 feature feature NOUN bracis-28396 285 37 selection selection NOUN bracis-28396 285 38 for for ADP bracis-28396 285 39 non non ADJ bracis-28396 285 40 - - ADJ bracis-28396 285 41 technical technical ADJ bracis-28396 285 42 losses loss NOUN bracis-28396 285 43 detection detection NOUN bracis-28396 285 44 . . PUNCT bracis-28396 286 1 in in ADP bracis-28396 286 2 : : PUNCT bracis-28396 286 3 tavares tavare NOUN bracis-28396 286 4 , , PUNCT bracis-28396 286 5 j.m.r.s j.m.r.s PROPN bracis-28396 286 6 . . PROPN bracis-28396 286 7 , , PUNCT bracis-28396 286 8 natal natal PROPN bracis-28396 286 9 jorge jorge NOUN bracis-28396 286 10 , , PUNCT bracis-28396 286 11 r.m r.m PROPN bracis-28396 286 12 . . PROPN bracis-28396 286 13 ( ( PUNCT bracis-28396 286 14 eds ed NOUN bracis-28396 286 15 . . PUNCT bracis-28396 286 16 ) ) PUNCT bracis-28396 286 17 vipimage vipimage NOUN bracis-28396 286 18 2019 2019 NUM bracis-28396 286 19 . . PUNCT bracis-28396 287 1 lncvb lncvb PROPN bracis-28396 287 2 , , PUNCT bracis-28396 287 3 vol vol NOUN bracis-28396 287 4 . . PROPN bracis-28396 287 5 34 34 NUM bracis-28396 287 6 , , PUNCT bracis-28396 287 7 pp pp ADJ bracis-28396 287 8 . . PUNCT bracis-28396 288 1 141–156 141–156 NUM bracis-28396 288 2 . . PUNCT bracis-28396 288 3 springer springer NOUN bracis-28396 288 4 , , PUNCT bracis-28396 288 5 cham cham PROPN bracis-28396 288 6 ( ( PUNCT bracis-28396 288 7 2019 2019 NUM bracis-28396 288 8 ) ) PUNCT bracis-28396 288 9 . . PUNCT bracis-28396 289 1 https://doi.org/10.1007/978-3-030-32040-9_16 https://doi.org/10.1007/978-3-030-32040-9_16 NOUN bracis-28396 289 2 chapter chapter NOUN bracis-28396 289 3     SPACE bracis-28396 289 4 google google PROPN bracis-28396 289 5 scholar scholar NOUN bracis-28396 289 6     SPACE bracis-28396 289 7 pham pham PROPN bracis-28396 289 8 , , PUNCT bracis-28396 289 9 d.t d.t PROPN bracis-28396 289 10 . . PROPN bracis-28396 289 11 , , PUNCT bracis-28396 289 12 alcock alcock PROPN bracis-28396 289 13 , , PUNCT bracis-28396 289 14 r.j r.j PROPN bracis-28396 289 15 . . PROPN bracis-28396 289 16 : : PUNCT bracis-28396 289 17 automatic automatic ADJ bracis-28396 289 18 detection detection NOUN bracis-28396 289 19 of of ADP bracis-28396 289 20 defects defect NOUN bracis-28396 289 21 on on ADP bracis-28396 289 22 birch birch NOUN bracis-28396 289 23 wood wood NOUN bracis-28396 289 24 boards board NOUN bracis-28396 289 25 . . PUNCT bracis-28396 290 1 proc proc NOUN bracis-28396 290 2 . . PUNCT bracis-28396 291 1 inst inst PROPN bracis-28396 291 2 . . PROPN bracis-28396 291 3 mech mech PROPN bracis-28396 291 4 . . PUNCT bracis-28396 292 1 eng eng PROPN bracis-28396 292 2 . . PUNCT bracis-28396 293 1 part part PROPN bracis-28396 293 2 e e PROPN bracis-28396 293 3 j. j. PROPN bracis-28396 293 4 process process PROPN bracis-28396 293 5 mech mech PROPN bracis-28396 293 6 . . PUNCT bracis-28396 294 1 eng eng PROPN bracis-28396 294 2 . . PROPN bracis-28396 294 3 210(1 210(1 NUM bracis-28396 294 4 ) ) PUNCT bracis-28396 294 5 , , PUNCT bracis-28396 294 6 45–52 45–52 NUM bracis-28396 294 7 ( ( PUNCT bracis-28396 294 8 1996 1996 NUM bracis-28396 294 9 ) ) PUNCT bracis-28396 294 10 . . PUNCT bracis-28396 295 1 https://doi.org/10.1243/0954408991529852 https://doi.org/10.1243/0954408991529852 PROPN bracis-28396 295 2 article article NOUN bracis-28396 295 3     SPACE bracis-28396 295 4 google google PROPN bracis-28396 295 5 scholar scholar NOUN bracis-28396 295 6     SPACE bracis-28396 295 7 qi qi PROPN bracis-28396 295 8 , , PUNCT bracis-28396 295 9 c. c. PROPN bracis-28396 295 10 , , PUNCT bracis-28396 295 11 fourie fourie PROPN bracis-28396 295 12 , , PUNCT bracis-28396 295 13 a. a. PROPN bracis-28396 295 14 , , PUNCT bracis-28396 295 15 chen chen PROPN bracis-28396 295 16 , , PUNCT bracis-28396 295 17 q. q. PROPN bracis-28396 295 18 : : PUNCT bracis-28396 295 19 neural neural ADJ bracis-28396 295 20 network network NOUN bracis-28396 295 21 and and CCONJ bracis-28396 295 22 particle particle NOUN bracis-28396 295 23 swarm swarm NOUN bracis-28396 295 24 optimization optimization NOUN bracis-28396 295 25 for for ADP bracis-28396 295 26 predicting predict VERB bracis-28396 295 27 the the DET bracis-28396 295 28 unconfined unconfined ADJ bracis-28396 295 29 compressive compressive ADJ bracis-28396 295 30 strength strength NOUN bracis-28396 295 31 of of ADP bracis-28396 295 32 cemented cement VERB bracis-28396 295 33 paste paste NOUN bracis-28396 295 34 backfill backfill NOUN bracis-28396 295 35 . . PUNCT bracis-28396 296 1 constr constr NOUN bracis-28396 296 2 . . PUNCT bracis-28396 297 1 build build VERB bracis-28396 297 2 . . PUNCT bracis-28396 298 1 mater mater NOUN bracis-28396 298 2 . . PROPN bracis-28396 298 3 159 159 NUM bracis-28396 298 4 , , PUNCT bracis-28396 298 5 473–478 473–478 NUM bracis-28396 298 6 ( ( PUNCT bracis-28396 298 7 2018 2018 NUM bracis-28396 298 8 ) ) PUNCT bracis-28396 298 9 . . PUNCT bracis-28396 299 1 https://doi.org/10.1016/j.conbuildmat.2017.11.006 https://doi.org/10.1016/j.conbuildmat.2017.11.006 ADJ bracis-28396 299 2 article article NOUN bracis-28396 299 3     SPACE bracis-28396 299 4 google google PROPN bracis-28396 299 5 scholar scholar NOUN bracis-28396 299 6     SPACE bracis-28396 299 7 r r NOUN bracis-28396 299 8 core core NOUN bracis-28396 299 9 team team NOUN bracis-28396 299 10 : : PUNCT bracis-28396 299 11 a a DET bracis-28396 299 12 language language NOUN bracis-28396 299 13 and and CCONJ bracis-28396 299 14 environment environment NOUN bracis-28396 299 15 for for ADP bracis-28396 299 16 statistical statistical ADJ bracis-28396 299 17 computing computing NOUN bracis-28396 299 18 . . PUNCT bracis-28396 300 1 r r NOUN bracis-28396 300 2 foundation foundation NOUN bracis-28396 300 3 for for ADP bracis-28396 300 4 statistical statistical ADJ bracis-28396 300 5 computing computing NOUN bracis-28396 300 6 , , PUNCT bracis-28396 300 7 vienna vienna PROPN bracis-28396 300 8 , , PUNCT bracis-28396 300 9 austria austria PROPN bracis-28396 300 10 ( ( PUNCT bracis-28396 300 11 2014 2014 NUM bracis-28396 300 12 ) ) PUNCT bracis-28396 300 13 google google NOUN bracis-28396 300 14 scholar scholar NOUN bracis-28396 300 15     SPACE bracis-28396 300 16 roder roder NOUN bracis-28396 300 17 , , PUNCT bracis-28396 300 18 m. m. NOUN bracis-28396 300 19 , , PUNCT bracis-28396 300 20 passos passos PROPN bracis-28396 300 21 , , PUNCT bracis-28396 300 22 l.a l.a PROPN bracis-28396 300 23 . . PROPN bracis-28396 300 24 , , PUNCT bracis-28396 300 25 de de PROPN bracis-28396 300 26 rosa rosa PROPN bracis-28396 300 27 , , PUNCT bracis-28396 300 28 g.h g.h PROPN bracis-28396 300 29 . . PROPN bracis-28396 300 30 , , PUNCT bracis-28396 300 31 de de X bracis-28396 300 32 albuquerque albuquerque NOUN bracis-28396 300 33 , , PUNCT bracis-28396 300 34 v.h.c v.h.c ADJ bracis-28396 300 35 . . PROPN bracis-28396 300 36 , , PUNCT bracis-28396 300 37 papa papa PROPN bracis-28396 300 38 , , PUNCT bracis-28396 300 39 j.p j.p PROPN bracis-28396 300 40 . . PROPN bracis-28396 300 41 : : PUNCT bracis-28396 300 42 reinforcing reinforce VERB bracis-28396 300 43 learning learn VERB bracis-28396 300 44 in in ADP bracis-28396 300 45 deep deep ADJ bracis-28396 300 46 belief belief NOUN bracis-28396 300 47 networks network NOUN bracis-28396 300 48 through through ADP bracis-28396 300 49 nature nature NOUN bracis-28396 300 50 - - PUNCT bracis-28396 300 51 inspired inspire VERB bracis-28396 300 52 optimization optimization NOUN bracis-28396 300 53 . . PUNCT bracis-28396 301 1 appl appl PROPN bracis-28396 301 2 . . PUNCT bracis-28396 301 3 soft soft ADJ bracis-28396 301 4 comput comput NOUN bracis-28396 301 5 . . PUNCT bracis-28396 302 1 108 108 NUM bracis-28396 302 2 , , PUNCT bracis-28396 302 3 107466 107466 NUM bracis-28396 302 4 ( ( PUNCT bracis-28396 302 5 2021 2021 NUM bracis-28396 302 6 ) ) PUNCT bracis-28396 302 7 article article NOUN bracis-28396 302 8     SPACE bracis-28396 302 9 google google PROPN bracis-28396 302 10 scholar scholar NOUN bracis-28396 302 11     SPACE bracis-28396 302 12 roder roder NOUN bracis-28396 302 13 , , PUNCT bracis-28396 302 14 m. m. NOUN bracis-28396 302 15 , , PUNCT bracis-28396 302 16 de de PROPN bracis-28396 302 17 rosa rosa PROPN bracis-28396 302 18 , , PUNCT bracis-28396 302 19 g.h g.h PROPN bracis-28396 302 20 . . PROPN bracis-28396 302 21 , , PUNCT bracis-28396 302 22 passos passos PROPN bracis-28396 302 23 , , PUNCT bracis-28396 302 24 l.a l.a PROPN bracis-28396 302 25 . . PROPN bracis-28396 302 26 , , PUNCT bracis-28396 302 27 papa papa PROPN bracis-28396 302 28 , , PUNCT bracis-28396 302 29 j.p j.p PROPN bracis-28396 302 30 . . PROPN bracis-28396 302 31 , , PUNCT bracis-28396 302 32 rossi rossi PROPN bracis-28396 302 33 , , PUNCT bracis-28396 302 34 a.l.d a.l.d NOUN bracis-28396 302 35 . . PUNCT bracis-28396 302 36 : : PUNCT bracis-28396 303 1 harnessing harness VERB bracis-28396 303 2 particle particle NOUN bracis-28396 303 3 swarm swarm NOUN bracis-28396 303 4 optimization optimization NOUN bracis-28396 303 5 through through ADP bracis-28396 303 6 relativistic relativistic ADJ bracis-28396 303 7 velocity velocity NOUN bracis-28396 303 8 . . PUNCT bracis-28396 304 1 in in ADP bracis-28396 304 2 : : PUNCT bracis-28396 304 3 2020 2020 NUM bracis-28396 304 4 ieee ieee PROPN bracis-28396 304 5 congress congress PROPN bracis-28396 304 6 on on ADP bracis-28396 304 7 evolutionary evolutionary ADJ bracis-28396 304 8 computation computation NOUN bracis-28396 304 9 ( ( PUNCT bracis-28396 304 10 cec cec PROPN bracis-28396 304 11 ) ) PUNCT bracis-28396 304 12 , , PUNCT bracis-28396 304 13 pp pp PROPN bracis-28396 304 14 . . PUNCT bracis-28396 305 1 1–8 1–8 X bracis-28396 305 2 . . PUNCT bracis-28396 305 3 ieee ieee PROPN bracis-28396 305 4 ( ( PUNCT bracis-28396 305 5 2020 2020 NUM bracis-28396 305 6 ) ) PUNCT bracis-28396 305 7 google google NOUN bracis-28396 305 8 scholar scholar NOUN bracis-28396 305 9     SPACE bracis-28396 305 10 roder roder NOUN bracis-28396 305 11 , , PUNCT bracis-28396 305 12 m. m. NOUN bracis-28396 305 13 , , PUNCT bracis-28396 305 14 rossi rossi ADJ bracis-28396 305 15 , , PUNCT bracis-28396 305 16 a.l.d a.l.d NOUN bracis-28396 305 17 . . PUNCT bracis-28396 305 18 , , PUNCT bracis-28396 305 19 de de PROPN bracis-28396 305 20     SPACE bracis-28396 305 21 oliveira oliveira PROPN bracis-28396 305 22     SPACE bracis-28396 305 23 affonso affonso PROPN bracis-28396 305 24 , , PUNCT bracis-28396 305 25 c. c. NOUN bracis-28396 305 26 : : PUNCT bracis-28396 305 27 boosting boost VERB bracis-28396 305 28 machine machine NOUN bracis-28396 305 29 learning learn VERB bracis-28396 305 30 techniques technique NOUN bracis-28396 305 31 for for ADP bracis-28396 305 32 wood wood NOUN bracis-28396 305 33 quality quality NOUN bracis-28396 305 34 classification classification NOUN bracis-28396 305 35 by by ADP bracis-28396 305 36 particle particle NOUN bracis-28396 305 37 swarm swarm NOUN bracis-28396 305 38 optimization optimization NOUN bracis-28396 305 39 . . PUNCT bracis-28396 306 1 in in ADP bracis-28396 306 2 : : PUNCT bracis-28396 306 3 encontro encontro ADJ bracis-28396 306 4 nacional nacional ADJ bracis-28396 306 5 de de X bracis-28396 306 6 inteligência inteligência PROPN bracis-28396 306 7 artificial artificial PROPN bracis-28396 306 8 e e PROPN bracis-28396 306 9 computacional computacional PROPN bracis-28396 306 10 . . PUNCT bracis-28396 307 1 sociedade sociedade PROPN bracis-28396 307 2 brasileira brasileira PROPN bracis-28396 307 3 de de X bracis-28396 307 4 computação computação X bracis-28396 307 5 ( ( PUNCT bracis-28396 307 6 2017 2017 NUM bracis-28396 307 7 ) ) PUNCT bracis-28396 307 8 google google PROPN bracis-28396 307 9 scholar scholar NOUN bracis-28396 307 10     SPACE bracis-28396 307 11 rodrigues rodrigue NOUN bracis-28396 307 12 , , PUNCT bracis-28396 307 13 d. d. PROPN bracis-28396 307 14 , , PUNCT bracis-28396 307 15 de de PROPN bracis-28396 307 16 rosa rosa PROPN bracis-28396 307 17 , , PUNCT bracis-28396 307 18 g.h g.h PROPN bracis-28396 307 19 . . PROPN bracis-28396 307 20 , , PUNCT bracis-28396 307 21 passos passos PROPN bracis-28396 307 22 , , PUNCT bracis-28396 307 23 l.a l.a PROPN bracis-28396 307 24 . . PROPN bracis-28396 307 25 , , PUNCT bracis-28396 307 26 papa papa PROPN bracis-28396 307 27 , , PUNCT bracis-28396 307 28 j.p j.p PROPN bracis-28396 307 29 . . PROPN bracis-28396 307 30 : : PUNCT bracis-28396 308 1 adaptive adaptive PROPN bracis-28396 308 2 improved improve VERB bracis-28396 308 3 flower flower NOUN bracis-28396 308 4 pollination pollination NOUN bracis-28396 308 5 algorithm algorithm NOUN bracis-28396 308 6 for for ADP bracis-28396 308 7 global global ADJ bracis-28396 308 8 optimization optimization NOUN bracis-28396 308 9 . . PUNCT bracis-28396 309 1 in in ADP bracis-28396 309 2 : : PUNCT bracis-28396 309 3 yang yang PROPN bracis-28396 309 4 , , PUNCT bracis-28396 309 5 x.-s x.-s PROPN bracis-28396 309 6 . . PROPN bracis-28396 309 7 , , PUNCT bracis-28396 309 8 he he PRON bracis-28396 309 9 , , PUNCT bracis-28396 309 10 x.-s x.-s PROPN bracis-28396 309 11 . . PUNCT bracis-28396 310 1 ( ( PUNCT bracis-28396 310 2 eds ed NOUN bracis-28396 310 3 . . PUNCT bracis-28396 310 4 ) ) PUNCT bracis-28396 310 5 nature nature NOUN bracis-28396 310 6 - - PUNCT bracis-28396 310 7 inspired inspire VERB bracis-28396 310 8 computation computation NOUN bracis-28396 310 9 in in ADP bracis-28396 310 10 data datum NOUN bracis-28396 310 11 mining mining NOUN bracis-28396 310 12 and and CCONJ bracis-28396 310 13 machine machine NOUN bracis-28396 310 14 learning learning NOUN bracis-28396 310 15 . . PUNCT bracis-28396 311 1 sci sci PROPN bracis-28396 311 2 , , PUNCT bracis-28396 311 3 vol vol NOUN bracis-28396 311 4 . . PROPN bracis-28396 311 5 855 855 NUM bracis-28396 311 6 , , PUNCT bracis-28396 311 7 pp pp ADJ bracis-28396 311 8 . . PUNCT bracis-28396 312 1 1–21 1–21 PROPN bracis-28396 312 2 . . PUNCT bracis-28396 312 3 springer springer NOUN bracis-28396 312 4 , , PUNCT bracis-28396 312 5 cham cham PROPN bracis-28396 312 6 ( ( PUNCT bracis-28396 312 7 2020 2020 NUM bracis-28396 312 8 ) ) PUNCT bracis-28396 312 9 . . PUNCT bracis-28396 313 1 https://doi.org/10.1007/978-3-030-28553-1_1 https://doi.org/10.1007/978-3-030-28553-1_1 PUNCT bracis-28396 313 2 chapter chapter NOUN bracis-28396 313 3     SPACE bracis-28396 313 4 google google PROPN bracis-28396 313 5 scholar scholar NOUN bracis-28396 313 6     SPACE bracis-28396 313 7 shi shi PROPN bracis-28396 313 8 , , PUNCT bracis-28396 313 9 y. y. PROPN bracis-28396 313 10 , , PUNCT bracis-28396 313 11 eberhart eberhart PROPN bracis-28396 313 12 , , PUNCT bracis-28396 313 13 r. r. PROPN bracis-28396 313 14 : : PUNCT bracis-28396 313 15 a a DET bracis-28396 313 16 modified modify VERB bracis-28396 313 17 particle particle NOUN bracis-28396 313 18 swarm swarm NOUN bracis-28396 313 19 optimizer optimizer NOUN bracis-28396 313 20 . . PUNCT bracis-28396 314 1 in in ADP bracis-28396 314 2 : : PUNCT bracis-28396 314 3 1998 1998 NUM bracis-28396 314 4 ieee ieee PROPN bracis-28396 314 5 international international PROPN bracis-28396 314 6 conference conference NOUN bracis-28396 314 7 on on ADP bracis-28396 314 8 evolutionary evolutionary ADJ bracis-28396 314 9 computation computation NOUN bracis-28396 314 10 proceedings proceeding NOUN bracis-28396 314 11 . . PUNCT bracis-28396 315 1 ieee ieee PROPN bracis-28396 315 2 world world PROPN bracis-28396 315 3 congress congress PROPN bracis-28396 315 4 on on ADP bracis-28396 315 5 computational computational ADJ bracis-28396 315 6 intelligence intelligence NOUN bracis-28396 315 7 ( ( PUNCT bracis-28396 315 8 cat cat NOUN bracis-28396 315 9 . . PUNCT bracis-28396 316 1 no no INTJ bracis-28396 316 2 . . NOUN bracis-28396 317 1 98th8360 98th8360 NUM bracis-28396 317 2 ) ) PUNCT bracis-28396 317 3 , , PUNCT bracis-28396 317 4 pp pp ADJ bracis-28396 317 5 . . PUNCT bracis-28396 318 1 69–73 69–73 NUM bracis-28396 318 2 , , PUNCT bracis-28396 318 3 may may PROPN bracis-28396 318 4 1998 1998 NUM bracis-28396 318 5 . . PUNCT bracis-28396 319 1 https://doi.org/10.1109/icec.1998.699146 https://doi.org/10.1109/icec.1998.699146 PROPN bracis-28396 319 2 tiryaki tiryaki ADJ bracis-28396 319 3 , , PUNCT bracis-28396 319 4 s. s. PROPN bracis-28396 319 5 , , PUNCT bracis-28396 319 6 malkoçoğlu malkoçoğlu PROPN bracis-28396 319 7 , , PUNCT bracis-28396 319 8 a. a. NOUN bracis-28396 319 9 , , PUNCT bracis-28396 319 10 özşahin özşahin NOUN bracis-28396 319 11 , , PUNCT bracis-28396 319 12 ş ş X bracis-28396 319 13 . . PUNCT bracis-28396 319 14 : : PUNCT bracis-28396 319 15 using use VERB bracis-28396 319 16 artificial artificial ADJ bracis-28396 319 17 neural neural ADJ bracis-28396 319 18 networks network NOUN bracis-28396 319 19 for for ADP bracis-28396 319 20 modeling model VERB bracis-28396 319 21 surface surface NOUN bracis-28396 319 22 roughness roughness NOUN bracis-28396 319 23 of of ADP bracis-28396 319 24 wood wood NOUN bracis-28396 319 25 in in ADP bracis-28396 319 26 machining machining NOUN bracis-28396 319 27 process process NOUN bracis-28396 319 28 . . PUNCT bracis-28396 320 1 constr constr NOUN bracis-28396 320 2 . . PUNCT bracis-28396 321 1 build build VERB bracis-28396 321 2 . . PUNCT bracis-28396 322 1 mater mater NOUN bracis-28396 322 2 . . PUNCT bracis-28396 323 1 66 66 NUM bracis-28396 323 2 , , PUNCT bracis-28396 323 3 329–335 329–335 NUM bracis-28396 323 4 ( ( PUNCT bracis-28396 323 5 2014 2014 NUM bracis-28396 323 6 ) ) PUNCT bracis-28396 323 7 . . PUNCT bracis-28396 324 1 https://doi.org/10.1016/j.conbuildmat.2014.05.098 https://doi.org/10.1016/j.conbuildmat.2014.05.098 NOUN bracis-28396 324 2 vieira vieira PROPN bracis-28396 324 3 , , PUNCT bracis-28396 324 4 f.h.a f.h.a NOUN bracis-28396 324 5 . . PUNCT bracis-28396 324 6 : : PUNCT bracis-28396 325 1 image image NOUN bracis-28396 325 2 processing process VERB bracis-28396 325 3 through through ADP bracis-28396 325 4 machine machine NOUN bracis-28396 325 5 learning learning NOUN bracis-28396 325 6 for for ADP bracis-28396 325 7 wood wood NOUN bracis-28396 325 8 quality quality NOUN bracis-28396 325 9 classification classification NOUN bracis-28396 325 10 . . PUNCT bracis-28396 326 1 ph.d ph.d PROPN bracis-28396 326 2 . . PUNCT bracis-28396 327 1 thesis thesis PROPN bracis-28396 327 2 , , PUNCT bracis-28396 327 3 faculdade faculdade PROPN bracis-28396 327 4 de de X bracis-28396 327 5 engenharia engenharia X bracis-28396 327 6 de de X bracis-28396 327 7 guaratinguetá guaratinguetá X bracis-28396 327 8 ( ( PUNCT bracis-28396 327 9 feg feg NOUN bracis-28396 327 10 ) ) PUNCT bracis-28396 327 11 , , PUNCT bracis-28396 327 12 unesp unesp NOUN bracis-28396 327 13 ( ( PUNCT bracis-28396 327 14 2016 2016 NUM bracis-28396 327 15 ) ) PUNCT bracis-28396 327 16 google google PROPN bracis-28396 327 17 scholar scholar NOUN bracis-28396 327 18     SPACE bracis-28396 327 19 wilcoxon wilcoxon ADJ bracis-28396 327 20 , , PUNCT bracis-28396 327 21 f. f. PROPN bracis-28396 327 22 : : PUNCT bracis-28396 327 23 individual individual ADJ bracis-28396 327 24 comparisons comparison NOUN bracis-28396 327 25 by by ADP bracis-28396 327 26 ranking rank VERB bracis-28396 327 27 methods method NOUN bracis-28396 327 28 . . PUNCT bracis-28396 328 1 biometrics biometric NOUN bracis-28396 328 2 bull bull PROPN bracis-28396 328 3 . . PUNCT bracis-28396 329 1 1(6 1(6 NUM bracis-28396 329 2 ) ) PUNCT bracis-28396 329 3 , , PUNCT bracis-28396 329 4 80–83 80–83 NUM bracis-28396 329 5 ( ( PUNCT bracis-28396 329 6 1945 1945 NUM bracis-28396 329 7 ) ) PUNCT bracis-28396 329 8 article article NOUN bracis-28396 329 9     SPACE bracis-28396 329 10 google google PROPN bracis-28396 329 11 scholar scholar NOUN bracis-28396 329 12     SPACE bracis-28396 329 13 download download NOUN bracis-28396 329 14 references reference NOUN bracis-28396 329 15 acknowledgments acknowledgment NOUN bracis-28396 329 16 the the DET bracis-28396 329 17 authors author NOUN bracis-28396 329 18 are be AUX bracis-28396 329 19 grateful grateful ADJ bracis-28396 329 20 to to ADP bracis-28396 329 21 fapesp fapesp ADJ bracis-28396 329 22 grants grant NOUN bracis-28396 329 23 # # SYM bracis-28396 329 24 2016/06538 2016/06538 NUM bracis-28396 329 25 - - SYM bracis-28396 329 26 0 0 NUM bracis-28396 329 27 , , PUNCT bracis-28396 330 1 # # SYM bracis-28396 330 2 2018/02822 2018/02822 NUM bracis-28396 330 3 - - SYM bracis-28396 330 4 1 1 NUM bracis-28396 330 5 , , PUNCT bracis-28396 330 6 # # SYM bracis-28396 330 7 2019/07825 2019/07825 NUM bracis-28396 330 8 - - SYM bracis-28396 330 9 1 1 NUM bracis-28396 330 10 , , PUNCT bracis-28396 330 11 and and CCONJ bracis-28396 330 12 # # SYM bracis-28396 330 13 2023/10823 2023/10823 NUM bracis-28396 330 14 - - SYM bracis-28396 330 15 6 6 NUM bracis-28396 330 16 author author NOUN bracis-28396 330 17 information information NOUN bracis-28396 330 18 authors author NOUN bracis-28396 330 19 and and CCONJ bracis-28396 330 20 affiliations affiliation NOUN bracis-28396 330 21 department department PROPN bracis-28396 330 22 of of ADP bracis-28396 330 23 computing computing PROPN bracis-28396 330 24 , , PUNCT bracis-28396 330 25 são são PROPN bracis-28396 330 26 paulo paulo PROPN bracis-28396 330 27 state state PROPN bracis-28396 330 28 university university PROPN bracis-28396 330 29 , , PUNCT bracis-28396 330 30 av av PROPN bracis-28396 330 31 . . PUNCT bracis-28396 330 32 eng eng PROPN bracis-28396 330 33 . . PROPN bracis-28396 330 34 luiz luiz PROPN bracis-28396 330 35 edmundo edmundo PROPN bracis-28396 330 36 carrijo carrijo PROPN bracis-28396 330 37 coube coube PROPN bracis-28396 330 38 , , PUNCT bracis-28396 330 39 14 14 NUM bracis-28396 330 40 - - SYM bracis-28396 330 41 01 01 NUM bracis-28396 330 42 , , PUNCT bracis-28396 330 43 bauru bauru PROPN bracis-28396 330 44 , , PUNCT bracis-28396 330 45 17033 17033 NUM bracis-28396 330 46 - - SYM bracis-28396 330 47 360 360 NUM bracis-28396 330 48 , , PUNCT bracis-28396 330 49 brazil brazil PROPN bracis-28396 330 50 mateus mateus PROPN bracis-28396 330 51 roder roder PROPN bracis-28396 330 52 , , PUNCT bracis-28396 330 53     SPACE bracis-28396 330 54 leandro leandro PROPN bracis-28396 330 55 aparecido aparecido PROPN bracis-28396 330 56 passos passos PROPN bracis-28396 330 57     SPACE bracis-28396 330 58 & & CCONJ bracis-28396 330 59     SPACE bracis-28396 330 60 joão joão PROPN bracis-28396 330 61 paulo paulo PROPN bracis-28396 330 62 papa papa PROPN bracis-28396 330 63 department department PROPN bracis-28396 330 64 of of ADP bracis-28396 330 65 production production PROPN bracis-28396 330 66 engineering engineering NOUN bracis-28396 330 67 , , PUNCT bracis-28396 330 68 paulo paulo PROPN bracis-28396 330 69 state state PROPN bracis-28396 330 70 university university PROPN bracis-28396 330 71 , , PUNCT bracis-28396 330 72 rua rua PROPN bracis-28396 330 73 geraldo geraldo PROPN bracis-28396 330 74 alckmin alckmin PROPN bracis-28396 330 75 , , PUNCT bracis-28396 330 76 519 519 NUM bracis-28396 330 77 vila vila PROPN bracis-28396 330 78 nossa nossa PROPN bracis-28396 330 79 sra sra PROPN bracis-28396 330 80 . . PROPN bracis-28396 330 81 de de PROPN bracis-28396 330 82 fatima fatima PROPN bracis-28396 330 83 , , PUNCT bracis-28396 330 84 itapeva itapeva PROPN bracis-28396 330 85 , , PUNCT bracis-28396 330 86 18409 18409 NUM bracis-28396 330 87 - - SYM bracis-28396 330 88 010 010 NUM bracis-28396 330 89 , , PUNCT bracis-28396 330 90 brazil brazil PROPN bracis-28396 330 91 andré andré VERB bracis-28396 330 92 luis luis PROPN bracis-28396 330 93 debiaso debiaso PROPN bracis-28396 330 94 rossi rossi PROPN bracis-28396 331 1 authors author NOUN bracis-28396 331 2 mateus mateus PROPN bracis-28396 331 3 roderview roderview PROPN bracis-28396 331 4 author author NOUN bracis-28396 331 5 publications publication VERB bracis-28396 331 6 search search NOUN bracis-28396 331 7 author author NOUN bracis-28396 331 8 on on ADP bracis-28396 331 9 : : PUNCT bracis-28396 331 10 pubmed pubmed PROPN bracis-28396 331 11     SPACE bracis-28396 331 12 google google PROPN bracis-28396 331 13 scholar scholar PROPN bracis-28396 331 14 leandro leandro PROPN bracis-28396 331 15 aparecido aparecido PROPN bracis-28396 331 16 passosview passosview PROPN bracis-28396 331 17 author author NOUN bracis-28396 331 18 publications publication NOUN bracis-28396 331 19 search search NOUN bracis-28396 331 20 author author NOUN bracis-28396 331 21 on on ADP bracis-28396 331 22 : : PUNCT bracis-28396 331 23 pubmed pubmed PROPN bracis-28396 331 24     SPACE bracis-28396 331 25 google google PROPN bracis-28396 331 26 scholar scholar PROPN bracis-28396 331 27 joão joão PROPN bracis-28396 331 28 paulo paulo PROPN bracis-28396 331 29 papaview papaview PROPN bracis-28396 331 30 author author NOUN bracis-28396 331 31 publications publication NOUN bracis-28396 331 32 search search NOUN bracis-28396 331 33 author author NOUN bracis-28396 331 34 on on ADP bracis-28396 331 35 : : PUNCT bracis-28396 331 36 pubmed pubmed PROPN bracis-28396 331 37     SPACE bracis-28396 331 38 google google PROPN bracis-28396 331 39 scholar scholar NOUN bracis-28396 331 40 andré andré VERB bracis-28396 331 41 luis luis PROPN bracis-28396 331 42 debiaso debiaso PROPN bracis-28396 331 43 rossiview rossiview NOUN bracis-28396 331 44 author author NOUN bracis-28396 331 45 publications publication NOUN bracis-28396 331 46 search search NOUN bracis-28396 331 47 author author NOUN bracis-28396 331 48 on on ADP bracis-28396 331 49 : : PUNCT bracis-28396 331 50 pubmed pubmed PROPN bracis-28396 331 51     SPACE bracis-28396 331 52 google google PROPN bracis-28396 331 53 scholar scholar NOUN bracis-28396 331 54 corresponding correspond VERB bracis-28396 331 55 author author NOUN bracis-28396 331 56 correspondence correspondence NOUN bracis-28396 331 57 to to ADP bracis-28396 331 58 leandro leandro PROPN bracis-28396 331 59 aparecido aparecido PROPN bracis-28396 331 60 passos passos PROPN bracis-28396 331 61 . . PUNCT bracis-28396 332 1 editor editor NOUN bracis-28396 332 2 information information NOUN bracis-28396 332 3 editors editor NOUN bracis-28396 332 4 and and CCONJ bracis-28396 332 5 affiliations affiliation NOUN bracis-28396 332 6 federal federal PROPN bracis-28396 332 7 university university PROPN bracis-28396 332 8 of of ADP bracis-28396 332 9 são são PROPN bracis-28396 332 10 carlos carlos PROPN bracis-28396 332 11 , , PUNCT bracis-28396 332 12 são são PROPN bracis-28396 332 13 carlos carlos PROPN bracis-28396 332 14 , , PUNCT bracis-28396 332 15 brazil brazil PROPN bracis-28396 332 16 murilo murilo PROPN bracis-28396 332 17 c. c. PROPN bracis-28396 332 18 naldi naldi PROPN bracis-28396 332 19 centro centro PROPN bracis-28396 332 20 universitario universitario PROPN bracis-28396 332 21 da da PROPN bracis-28396 332 22 fei fei PROPN bracis-28396 332 23 , , PUNCT bracis-28396 332 24 são são PROPN bracis-28396 332 25 bernardo bernardo PROPN bracis-28396 332 26 do do AUX bracis-28396 332 27 campo campo PROPN bracis-28396 332 28 , , PUNCT bracis-28396 332 29 brazil brazil PROPN bracis-28396 332 30 reinaldo reinaldo PROPN bracis-28396 332 31 a. a. PROPN bracis-28396 332 32 c. c. PROPN bracis-28396 332 33 bianchi bianchi PROPN bracis-28396 332 34 rights right NOUN bracis-28396 332 35 and and CCONJ bracis-28396 332 36 permissions permission NOUN bracis-28396 332 37 reprints reprint NOUN bracis-28396 332 38 and and CCONJ bracis-28396 332 39 permissions permission VERB bracis-28396 332 40 copyright copyright NOUN bracis-28396 332 41 information information NOUN bracis-28396 332 42 © © ADP bracis-28396 332 43 2023 2023 NUM bracis-28396 332 44 the the DET bracis-28396 332 45 author(s author(s NOUN bracis-28396 332 46 ) ) PUNCT bracis-28396 332 47 , , PUNCT bracis-28396 332 48 under under ADP bracis-28396 332 49 exclusive exclusive ADJ bracis-28396 332 50 license license NOUN bracis-28396 332 51 to to ADP bracis-28396 332 52 springer springer NOUN bracis-28396 332 53 nature nature PROPN bracis-28396 332 54 switzerland switzerland PROPN bracis-28396 332 55 ag ag PROPN bracis-28396 332 56 about about ADP bracis-28396 332 57 this this DET bracis-28396 332 58 paper paper NOUN bracis-28396 332 59 cite cite VERB bracis-28396 332 60 this this DET bracis-28396 332 61 paper paper NOUN bracis-28396 332 62 roder roder NOUN bracis-28396 332 63 , , PUNCT bracis-28396 332 64 m. m. NOUN bracis-28396 332 65 , , PUNCT bracis-28396 332 66 passos passos PROPN bracis-28396 332 67 , , PUNCT bracis-28396 332 68 l.a l.a PROPN bracis-28396 332 69 . . PROPN bracis-28396 332 70 , , PUNCT bracis-28396 332 71 papa papa PROPN bracis-28396 332 72 , , PUNCT bracis-28396 332 73 j.p j.p PROPN bracis-28396 332 74 . . PROPN bracis-28396 332 75 , , PUNCT bracis-28396 332 76 rossi rossi PROPN bracis-28396 332 77 , , PUNCT bracis-28396 332 78 a.l.d a.l.d NOUN bracis-28396 332 79 . . PUNCT bracis-28396 333 1 ( ( PUNCT bracis-28396 333 2 2023 2023 NUM bracis-28396 333 3 ) ) PUNCT bracis-28396 333 4 . . PUNCT bracis-28396 334 1 feature feature NOUN bracis-28396 334 2 selection selection NOUN bracis-28396 334 3 and and CCONJ bracis-28396 334 4     SPACE bracis-28396 334 5 hyperparameter hyperparameter NOUN bracis-28396 334 6 fine fine ADV bracis-28396 334 7 - - PUNCT bracis-28396 334 8 tuning tuning NOUN bracis-28396 334 9 in in ADP bracis-28396 334 10     SPACE bracis-28396 334 11 artificial artificial ADJ bracis-28396 334 12 neural neural ADJ bracis-28396 334 13 networks network NOUN bracis-28396 334 14 for for ADP bracis-28396 334 15     SPACE bracis-28396 334 16 wood wood NOUN bracis-28396 334 17 quality quality NOUN bracis-28396 334 18 classification classification NOUN bracis-28396 334 19 . . PUNCT bracis-28396 335 1 in in ADP bracis-28396 335 2 : : PUNCT bracis-28396 335 3 naldi naldi PROPN bracis-28396 335 4 , , PUNCT bracis-28396 335 5 m.c m.c PROPN bracis-28396 335 6 . . PROPN bracis-28396 335 7 , , PUNCT bracis-28396 335 8 bianchi bianchi PROPN bracis-28396 335 9 , , PUNCT bracis-28396 335 10 r.a.c r.a.c ADP bracis-28396 335 11 . . PUNCT bracis-28396 335 12 ( ( PUNCT bracis-28396 335 13 eds ed NOUN bracis-28396 335 14 ) ) PUNCT bracis-28396 335 15 intelligent intelligent ADJ bracis-28396 335 16 systems system NOUN bracis-28396 335 17 . . PUNCT bracis-28396 336 1 bracis bracis PROPN bracis-28396 336 2 2023 2023 NUM bracis-28396 336 3 . . PUNCT bracis-28396 337 1 lecture lecture NOUN bracis-28396 337 2 notes note NOUN bracis-28396 337 3 in in ADP bracis-28396 337 4 computer computer NOUN bracis-28396 337 5 science science NOUN bracis-28396 337 6 ( ( PUNCT bracis-28396 337 7 ) ) PUNCT bracis-28396 337 8 , , PUNCT bracis-28396 337 9 vol vol NOUN bracis-28396 337 10 14196 14196 NUM bracis-28396 337 11 . . PUNCT bracis-28396 338 1 springer springer NOUN bracis-28396 338 2 , , PUNCT bracis-28396 338 3 cham cham PROPN bracis-28396 338 4 . . PUNCT bracis-28396 339 1 https://doi.org/10.1007/978-3-031-45389-2_22 https://doi.org/10.1007/978-3-031-45389-2_22 PROPN bracis-28396 339 2 download download NOUN bracis-28396 339 3 citation citation NOUN bracis-28396 339 4 .ris .ris PUNCT bracis-28396 340 1 .enw .enw PROPN bracis-28396 340 2 .bib .bib PUNCT bracis-28396 341 1 doi doi PROPN bracis-28396 341 2 : : PUNCT bracis-28396 341 3 https://doi.org/10.1007/978-3-031-45389-2_22 https://doi.org/10.1007/978-3-031-45389-2_22 VERB bracis-28396 341 4 published publish VERB bracis-28396 341 5 : : PUNCT bracis-28396 341 6 12 12 NUM bracis-28396 341 7 october october PROPN bracis-28396 341 8 2023 2023 NUM bracis-28396 341 9 publisher publisher NOUN bracis-28396 341 10 name name NOUN bracis-28396 341 11 : : PUNCT bracis-28396 341 12 springer springer NOUN bracis-28396 341 13 , , PUNCT bracis-28396 341 14 cham cham PROPN bracis-28396 341 15 print print PROPN bracis-28396 341 16 isbn isbn PROPN bracis-28396 341 17 : : PUNCT bracis-28396 341 18 978 978 NUM bracis-28396 341 19 - - SYM bracis-28396 341 20 3 3 NUM bracis-28396 341 21 - - PUNCT bracis-28396 341 22 031 031 NUM bracis-28396 341 23 - - PUNCT bracis-28396 341 24 45388 45388 NUM bracis-28396 341 25 - - SYM bracis-28396 341 26 5 5 NUM bracis-28396 341 27 online online ADJ bracis-28396 341 28 isbn isbn NOUN bracis-28396 341 29 : : PUNCT bracis-28396 341 30 978 978 NUM bracis-28396 341 31 - - SYM bracis-28396 341 32 3 3 NUM bracis-28396 341 33 - - PUNCT bracis-28396 341 34 031 031 NUM bracis-28396 341 35 - - PUNCT bracis-28396 341 36 45389 45389 NUM bracis-28396 341 37 - - SYM bracis-28396 341 38 2 2 NUM bracis-28396 341 39 ebook ebook NOUN bracis-28396 341 40 packages package NOUN bracis-28396 341 41 : : PUNCT bracis-28396 341 42 computer computer NOUN bracis-28396 341 43 sciencecomputer sciencecomputer NOUN bracis-28396 341 44 science science NOUN bracis-28396 341 45 ( ( PUNCT bracis-28396 341 46 r0 r0 NOUN bracis-28396 341 47 ) ) PUNCT bracis-28396 341 48 share share VERB bracis-28396 341 49 this this DET bracis-28396 341 50 paper paper NOUN bracis-28396 341 51 anyone anyone PRON bracis-28396 341 52 you you PRON bracis-28396 341 53 share share VERB bracis-28396 341 54 the the DET bracis-28396 341 55 following follow VERB bracis-28396 341 56 link link NOUN bracis-28396 341 57 with with ADP bracis-28396 341 58 will will AUX bracis-28396 341 59 be be AUX bracis-28396 341 60 able able ADJ bracis-28396 341 61 to to PART bracis-28396 341 62 read read VERB bracis-28396 341 63 this this DET bracis-28396 341 64 content content NOUN bracis-28396 341 65 : : PUNCT bracis-28396 341 66 get get VERB bracis-28396 341 67 shareable shareable ADJ bracis-28396 341 68 linksorry linksorry NOUN bracis-28396 341 69 , , PUNCT bracis-28396 341 70 a a DET bracis-28396 341 71 shareable shareable ADJ bracis-28396 341 72 link link NOUN bracis-28396 341 73 is be AUX bracis-28396 341 74 not not PART bracis-28396 341 75 currently currently ADV bracis-28396 341 76 available available ADJ bracis-28396 341 77 for for ADP bracis-28396 341 78 this this DET bracis-28396 341 79 article article NOUN bracis-28396 341 80 . . PUNCT bracis-28396 342 1 copy copy VERB bracis-28396 342 2 shareable shareable ADJ bracis-28396 342 3 link link NOUN bracis-28396 342 4 to to PART bracis-28396 342 5 clipboard clipboard NOUN bracis-28396 342 6 provided provide VERB bracis-28396 342 7 by by ADP bracis-28396 342 8 the the DET bracis-28396 342 9 springer springer NOUN bracis-28396 342 10 nature nature PROPN bracis-28396 342 11 sharedit sharedit PROPN bracis-28396 342 12 content content NOUN bracis-28396 342 13 - - PUNCT bracis-28396 342 14 sharing share VERB bracis-28396 342 15 initiative initiative NOUN bracis-28396 342 16 keywords keyword VERB bracis-28396 342 17 artificial artificial ADJ bracis-28396 342 18 neural neural ADJ bracis-28396 342 19 network network NOUN bracis-28396 342 20 optimization optimization NOUN bracis-28396 342 21 wood wood NOUN bracis-28396 342 22 quality quality NOUN bracis-28396 342 23 classification classification NOUN bracis-28396 342 24 sawmill sawmill NOUN bracis-28396 342 25 problem problem NOUN bracis-28396 342 26 hyperparameter hyperparameter NOUN bracis-28396 342 27 tuning tune VERB bracis-28396 342 28 feature feature NOUN bracis-28396 342 29 selection selection NOUN bracis-28396 342 30 publish publish NOUN bracis-28396 342 31 with with ADP bracis-28396 342 32 us us PROPN bracis-28396 342 33 policies policy NOUN bracis-28396 342 34 and and CCONJ bracis-28396 342 35 ethics ethic NOUN bracis-28396 342 36 profiles profile NOUN bracis-28396 342 37 leandro leandro PROPN bracis-28396 342 38 aparecido aparecido PROPN bracis-28396 342 39 passos passos PROPN bracis-28396 342 40 view view PROPN bracis-28396 342 41 author author NOUN bracis-28396 342 42 profile profile PROPN bracis-28396 342 43 joão joão PROPN bracis-28396 342 44 paulo paulo PROPN bracis-28396 342 45 papa papa PROPN bracis-28396 342 46 view view NOUN bracis-28396 342 47 author author NOUN bracis-28396 342 48 profile profile NOUN bracis-28396 342 49 search search NOUN bracis-28396 342 50 search search NOUN bracis-28396 342 51 by by ADP bracis-28396 342 52 keyword keyword NOUN bracis-28396 342 53 or or CCONJ bracis-28396 342 54 author author NOUN bracis-28396 342 55 search search NOUN bracis-28396 342 56 navigation navigation NOUN bracis-28396 342 57 find find VERB bracis-28396 342 58 a a DET bracis-28396 342 59 journal journal NOUN bracis-28396 342 60 publish publish VERB bracis-28396 342 61 with with ADP bracis-28396 342 62 us we PRON bracis-28396 342 63 track track VERB bracis-28396 342 64 your your PRON bracis-28396 342 65 research research NOUN bracis-28396 342 66 discover discover VERB bracis-28396 342 67 content content NOUN bracis-28396 342 68 journals journal NOUN bracis-28396 342 69 a a DET bracis-28396 342 70 - - PUNCT bracis-28396 342 71 z z NOUN bracis-28396 342 72 books book NOUN bracis-28396 342 73 a a DET bracis-28396 342 74 - - PUNCT bracis-28396 342 75 z z NOUN bracis-28396 342 76 publish publish NOUN bracis-28396 342 77 with with ADP bracis-28396 342 78 us us PROPN bracis-28396 342 79 journal journal PROPN bracis-28396 342 80 finder finder PROPN bracis-28396 342 81 publish publish VERB bracis-28396 342 82 your your PRON bracis-28396 342 83 research research NOUN bracis-28396 342 84 language language NOUN bracis-28396 342 85 editing edit VERB bracis-28396 342 86 open open ADJ bracis-28396 342 87 access access NOUN bracis-28396 342 88 publishing publishing NOUN bracis-28396 342 89 products product NOUN bracis-28396 342 90 and and CCONJ bracis-28396 342 91 services service NOUN bracis-28396 342 92 our our PRON bracis-28396 342 93 products product NOUN bracis-28396 342 94 librarians librarian VERB bracis-28396 342 95 societies society NOUN bracis-28396 342 96 partners partner NOUN bracis-28396 342 97 and and CCONJ bracis-28396 342 98 advertisers advertiser NOUN bracis-28396 342 99 our our PRON bracis-28396 342 100 brands brand NOUN bracis-28396 342 101 springer springer NOUN bracis-28396 342 102 nature nature PROPN bracis-28396 342 103 portfolio portfolio PROPN bracis-28396 342 104 bmc bmc PROPN bracis-28396 342 105 palgrave palgrave PROPN bracis-28396 342 106 macmillan macmillan PROPN bracis-28396 342 107 apress apress PROPN bracis-28396 342 108 discover discover VERB bracis-28396 342 109 your your PRON bracis-28396 342 110 privacy privacy NOUN bracis-28396 342 111 choices choice NOUN bracis-28396 342 112 / / SYM bracis-28396 342 113 manage manage NOUN bracis-28396 342 114 cookies cookie NOUN bracis-28396 342 115 your your PRON bracis-28396 342 116 us us PROPN bracis-28396 343 1 state state NOUN bracis-28396 343 2 privacy privacy NOUN bracis-28396 343 3 rights right NOUN bracis-28396 343 4 accessibility accessibility NOUN bracis-28396 343 5 statement statement NOUN bracis-28396 343 6 terms term NOUN bracis-28396 343 7 and and CCONJ bracis-28396 343 8 conditions condition NOUN bracis-28396 343 9 privacy privacy NOUN bracis-28396 343 10 policy policy NOUN bracis-28396 343 11 help help NOUN bracis-28396 343 12 and and CCONJ bracis-28396 343 13 support support VERB bracis-28396 343 14 legal legal ADJ bracis-28396 343 15 notice notice NOUN bracis-28396 343 16 cancel cancel VERB bracis-28396 343 17 contracts contract NOUN bracis-28396 343 18 here here ADV bracis-28396 343 19 129.74.145.123 129.74.145.123 NUM bracis-28396 343 20 hesburgh hesburgh PROPN bracis-28396 343 21 library library PROPN bracis-28396 343 22 er er INTJ bracis-28396 343 23 unit unit NOUN bracis-28396 343 24 ( ( PUNCT bracis-28396 343 25 3005732405 3005732405 NUM bracis-28396 343 26 ) ) PUNCT bracis-28396 343 27 northeast northeast ADJ bracis-28396 343 28 research research NOUN bracis-28396 343 29 libraries library NOUN bracis-28396 343 30 ( ( PUNCT bracis-28396 343 31 nerl nerl PROPN bracis-28396 343 32 ) ) PUNCT bracis-28396 343 33 ( ( PUNCT bracis-28396 343 34 8200828607 8200828607 NUM bracis-28396 343 35 ) ) PUNCT bracis-28396 343 36 nerl nerl VERB bracis-28396 343 37 ta ta X bracis-28396 343 38 account account NOUN bracis-28396 343 39 ( ( PUNCT bracis-28396 343 40 3006206169 3006206169 NUM bracis-28396 343 41 ) ) PUNCT bracis-28396 343 42 university university NOUN bracis-28396 343 43 of of ADP bracis-28396 343 44 notre notre PROPN bracis-28396 343 45 dame dame PROPN bracis-28396 343 46 hesburgh hesburgh PROPN bracis-28396 343 47 library library NOUN bracis-28396 343 48 ( ( PUNCT bracis-28396 343 49 3000184373 3000184373 NUM bracis-28396 343 50 ) ) PUNCT bracis-28396 344 1 © © ADP bracis-28396 344 2 2025 2025 NUM bracis-28396 344 3 springer springer NOUN bracis-28396 344 4 nature nature NOUN