The Community for Technology Leaders
Pattern Recognition, International Conference on (2004)
Cambridge UK
Aug. 23, 2004 to Aug. 26, 2004
ISSN: 1051-4651
ISBN: 0-7695-2128-2
pp: 582-585
Fei Wu , Tsinghua University, Beijing, China
Yonglei Zhou , Tsinghua University, Beijing, China
Changshui Zhang , Tsinghua University, Beijing, China
"Learning with side-information" is attracting more and more attention in machine learning problems. In this paper, we propose a general iterative framework for relevant linear feature extraction. It efficiently utilizes both the side-information and unlabeled data to enhance gradually algorithms' performance and robustness. Both good relevant feature extraction and reasonable similarity matrix estima-tion can be realized. Specifically, we adopt Relevant Component Analysis (RCA) under this framework and get the derived Iterative Self-Enhanced Relevant Component Analysis (ISERCA) algorithm. The experimental results on several data sets show that ISERCA outperforms RCA.

F. Wu, Y. Zhou and C. Zhang, "Relevant Linear Feature Extraction Using Side-information and Unlabeled Data," Pattern Recognition, International Conference on(ICPR), Cambridge UK, 2004, pp. 582-585.
94 ms
(Ver 3.3 (11022016))