For Full-Text PDF, please login, if you are a member of IEICE,|
or go to Pay Per View on menu list, if you are a nonmember of IEICE.
Improving the Accuracy of Least-Squares Probabilistic Classifiers
Makoto YAMADA Masashi SUGIYAMA Gordon WICHERN Jaak SIMM
IEICE TRANSACTIONS on Information and Systems
Publication Date: 2011/06/01
Online ISSN: 1745-1361
Print ISSN: 0916-8532
Type of Manuscript: LETTER
Category: Pattern Recognition
least-squares probabilistic classifier, kernel logistic regression, density ratio, PASCAL VOC 2010, freesound,
Full Text: PDF>>
The least-squares probabilistic classifier (LSPC) is a computationally-efficient alternative to kernel logistic regression. However, to assure its learned probabilities to be non-negative, LSPC involves a post-processing step of rounding up negative parameters to zero, which can unexpectedly influence classification performance. In order to mitigate 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 classification accuracy, while the computational advantage of the original LSPC remains unchanged.