GSML: A Unified Framework for Sparse Metric Learning

Huang Kaizhu, Ying Yiming, ICG Campbell

Research output: Chapter in Book/Report/Conference proceedingConference Contribution (Conference Proceeding)

37 Citations (Scopus)


There has been significant recent interest in sparse metric learning (SML) in which we simultaneously learn both a good distance metric and a low-dimensional representation. Unfortunately, the performance of existing sparse metric learning approaches is usually limited because the authors assumed certain problem relaxations or they target the SML objective indirectly. In this paper, we propose a Generalized Sparse Metric Learning method (GSML). This novel framework offers a unified view for understanding many of the popular sparse metric learning algorithms including the Sparse Metric Learning framework proposed in [15], the Large Margin Nearest Neighbor (LMNN) [21][22], and the D-ranking Vector Machine (D-ranking VM) [14]. Moreover, GSML also establishes a close relationship with the Pairwise Support Vector Machine [20]. Furthermore, the proposed framework is capable of extending many current non-sparse metric learning models such as Relevant Vector Machine (RCA) [4] and a state-of-the-art method proposed in [23] into their sparse versions. We present the detailed framework, provide theoretical justifications, build various connections with other models, and propose a practical iterative optimization method, making the framework both theoretically important and practically scalable for medium or large datasets. A series of experiments show that the proposed approach can outperform previous methods in terms of both test accuracy and dimension reduction, on six realworld benchmark datasets.
Translated title of the contributionGSML: A Unified Framework for Sparse Metric Learning
Original languageEnglish
Title of host publicationInternational Conference on Data Mining
Publication statusPublished - 2009

Bibliographical note

Name and Venue of Event: ICDM Miami, USA, 2009
Conference Proceedings/Title of Journal: Proceedings IEEE International Conference on Data Mining, ICDM 2009


Dive into the research topics of 'GSML: A Unified Framework for Sparse Metric Learning'. Together they form a unique fingerprint.

Cite this