Presentation 2010-09-05
Improving the Accuracy of Least-Squares Probabilistic Classifiers
Makoto YAMADA, Masashi SUGIYAMA, Gordon WICHERN, Jaak SIMM,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Least-Squares Probabilistic Classifier (LSPC) has been demonstrated to be a computationally-efficient and accurate classification method. However, since LSPC involves a post-processing step of rounding up its negative parameters to zero for assuring learned probabilities to be non-negative, its classification performance can be unexpectedly changed. In order to avoid this problem, we propose a simple alternative scheme that directly rounds up the classifier's negative outputs, not negative parameters. Through extensive experiments including real-world image classification and audio tagging tasks, we demonstrate that the proposed modification significantly improves the classification accuracy, while the computational advantage of the original LSPC is kept unchanged.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Least-Squares Probabilistic Classifier / Kernel Logistic Regression / Density Ratio / PASCAL VOC 2010 / Freesound
Paper # PRMU2010-60,IBISML2010-32
Date of Issue

Conference Information
Committee PRMU
Conference Date 2010/8/29(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Pattern Recognition and Media Understanding (PRMU)
Language ENG
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Improving the Accuracy of Least-Squares Probabilistic Classifiers
Sub Title (in English)
Keyword(1) Least-Squares Probabilistic Classifier
Keyword(2) Kernel Logistic Regression
Keyword(3) Density Ratio
Keyword(4) PASCAL VOC 2010
Keyword(5) Freesound
1st Author's Name Makoto YAMADA
1st Author's Affiliation Department of Computer Science, Tokyo Institute of Technology()
2nd Author's Name Masashi SUGIYAMA
2nd Author's Affiliation Department of Computer Science, Tokyo Institute of Technology
3rd Author's Name Gordon WICHERN
3rd Author's Affiliation MIT Lincoln Laboratory
4th Author's Name Jaak SIMM
4th Author's Affiliation Department of Computer Science, Tokyo Institute of Technology
Date 2010-09-05
Paper # PRMU2010-60,IBISML2010-32
Volume (vol) vol.110
Number (no) 187
Page pp.pp.-
#Pages 6
Date of Issue