Browse

A design framework for hierarchical ensemble of multiple feature extractors and multiple classifiers

DC Field Value Language
dc.contributor.authorKim, Kyounghoonen
dc.contributor.authorLin, Helin-
dc.contributor.authorChoi, Jin Young-
dc.contributor.authorChoi, Kiyoung-
dc.date.accessioned2017-04-19T00:45:58Z-
dc.date.available2017-12-05T11:19:05Z-
dc.date.issued2016-04-
dc.identifier.citationPattern Recognition, Vol.52, pp. 1-16-
dc.identifier.issn0031-3203-
dc.identifier.urihttps://hdl.handle.net/10371/116941-
dc.description.abstractIt is well-known that ensemble of classifiers can achieve higher accuracy compared to a single classifier system. This paper pays attention to ensemble systems consisting of multiple feature extractors and multiple classifiers (MFMC). However, MFMC increases the system complexity dramatically, leading to a highly complex combinatorial optimization problem. In order to overcome the complexity while exploiting the diversity of MFMC, we suggest in this paper a hierarchical ensemble of MFMC and its optimizing framework. By constructing local groups of feature extractors and classifiers and then combining them as a global group, the approach achieves a better scalability. Both reinforcement machine learning and Bayesian networks are adopted to enhance the accuracy. We apply the proposed method to vision based pedestrian detection and recognition of handwritten numerals. Experimental results show that the proposed framework outperforms the previous ensemble methods in terms of accuracy. (C) 2015 Elsevier Ltd. All rights reserved.en
dc.language.isoen-
dc.publisherELSEVIER SCI LTDen
dc.subjectA design framework for hierarchical ensemble of multiple feature extractors and multiple classifiersen
dc.subject자연과학en
dc.subjectEnsemble of detection systems-
dc.subjectMultiple feature extractors-
dc.subjectMultiple classifiers-
dc.subjectPedestrian detection-
dc.subjectReinforcement learning-
dc.subjectBayesian network-
dc.titleA design framework for hierarchical ensemble of multiple feature extractors and multiple classifiersen
dc.typeArticleen
dc.contributor.AlternativeAuthor김경훈-
dc.contributor.AlternativeAuthor이혜림-
dc.contributor.AlternativeAuthor최진영-
dc.contributor.AlternativeAuthor최기영-
dc.identifier.doi10.1016/j.patcog.2015.11.006-
dc.identifier.doi10.1016/j.patcog.2015.11.006-
dc.citation.journaltitlePattern Recognition-
dc.description.srndOAIID:RECH_ACHV_DSTSH_NO:T201622918-
dc.description.srndRECH_ACHV_FG:RR00200001-
dc.description.srndADJUST_YN:-
dc.description.srndEMP_ID:A004911-
dc.description.srndCITE_RATE:3.399-
dc.description.srndDEPT_NM:전기·정보공학부-
dc.description.srndEMAIL:jychoi@snu.ac.kr-
dc.description.srndSCOPUS_YN:Y-
dc.description.srndCONFIRM:Y-
dc.identifier.srndT201622918-
Appears in Collections:
College of Engineering/Engineering Practice School (공과대학/대학원)Dept. of Electrical and Computer Engineering (전기·정보공학부)Journal Papers (저널논문_전기·정보공학부)
Files in This Item:
There are no files associated with this item.
  • mendeley

Items in S-Space are protected by copyright, with all rights reserved, unless otherwise indicated.

Browse