IEICE TRANSACTIONS on Information
Improving the Accuracy of Least-Squares Probabilistic Classifiers
Summary :
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.
- Publication
- IEICE TRANSACTIONS on Information Vol.E94-D No.6 pp.1337-1340
- Publication Date
- 2011/06/01
- Publicized
- Online ISSN
- 1745-1361
- DOI
- 10.1587/transinf.E94.D.1337
- Type of Manuscript
- LETTER
- Category
- Pattern Recognition
Authors
Keyword
Latest Issue
Copyrights notice of machine-translated contents
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Cite this
Copy
Makoto YAMADA, Masashi SUGIYAMA, Gordon WICHERN, Jaak SIMM, "Improving the Accuracy of Least-Squares Probabilistic Classifiers" in IEICE TRANSACTIONS on Information,
vol. E94-D, no. 6, pp. 1337-1340, June 2011, doi: 10.1587/transinf.E94.D.1337.
Abstract: 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.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.E94.D.1337/_p
Copy
@ARTICLE{e94-d_6_1337,
author={Makoto YAMADA, Masashi SUGIYAMA, Gordon WICHERN, Jaak SIMM, },
journal={IEICE TRANSACTIONS on Information},
title={Improving the Accuracy of Least-Squares Probabilistic Classifiers},
year={2011},
volume={E94-D},
number={6},
pages={1337-1340},
abstract={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.},
keywords={},
doi={10.1587/transinf.E94.D.1337},
ISSN={1745-1361},
month={June},}
Copy
TY - JOUR
TI - Improving the Accuracy of Least-Squares Probabilistic Classifiers
T2 - IEICE TRANSACTIONS on Information
SP - 1337
EP - 1340
AU - Makoto YAMADA
AU - Masashi SUGIYAMA
AU - Gordon WICHERN
AU - Jaak SIMM
PY - 2011
DO - 10.1587/transinf.E94.D.1337
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E94-D
IS - 6
JA - IEICE TRANSACTIONS on Information
Y1 - June 2011
AB - 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.
ER -
English
