Towards Making High Dimensional Distance Metric Learning Practical

TL;DR

This paper introduces a dual random projection method for high-dimensional distance metric learning that avoids the suboptimality of traditional dimensionality reduction, combining efficiency with improved accuracy.

Contribution

The authors propose a novel dual random projection framework that effectively addresses the limitations of standard dimensionality reduction in high-dimensional DML.

Findings

The method achieves comparable or better accuracy than traditional approaches.

The approach is computationally efficient due to the use of random projections.

Theoretical analysis confirms the effectiveness of the proposed algorithm.

Abstract

In this work, we study distance metric learning (DML) for high dimensional data. A typical approach for DML with high dimensional data is to perform the dimensionality reduction first before learning the distance metric. The main shortcoming of this approach is that it may result in a suboptimal solution due to the subspace removed by the dimensionality reduction method. In this work, we present a dual random projection frame for DML with high dimensional data that explicitly addresses the limitation of dimensionality reduction for DML. The key idea is to first project all the data points into a low dimensional space by random projection, and compute the dual variables using the projected vectors. It then reconstructs the distance metric in the original space using the estimated dual variables. The proposed method, on one hand, enjoys the light computation of random projection, and on the other hand, alleviates the limitation of most dimensionality reduction methods. We verify both empirically and theoretically the effectiveness of the proposed algorithm for high dimensional DML.