Please use this identifier to cite or link to this item:

Kernel-based metric adaptation with pairwise constraints

Authors Chang, Hong
Yeung, Dit-Yan
Issue Date 2006
Source Lecture Notes in Artificial Intelligence , v. 3930, 2006, p. 721-730
Summary Many supervised and unsupervised learning algorithms depend on the choice of an appropriate distance metric. While metric learning for supervised learning tasks has a long history, extending it to learning tasks with weaker supervisory information has only been studied very recently. In particular, several methods have been proposed for semi-supervised metric learning based on pairwise (dis)similarity information. In this paper, we propose a kernel-based approach for nonlinear metric learning, which performs locally linear translation in the kernel-induced feature space. We formulate the metric learning problem as a kernel learning problem and solve it efficiently by kernel matrix adaptation. Experimental results based on synthetic and real-world data sets show that our approach is promising for semi-supervised metric learning.
ISSN 0302-9743
Rights The original publication is available at
Language English
Format Article
Access View full-text via Web of Science
View full-text via Scopus
Files in this item:
File Description Size Format
yeung.book20061.pdf 321549 B Adobe PDF