Generalized Leverage Scores: Geometric Interpretation and Applications

TL;DR

This paper extends leverage scores to relate matrix columns to arbitrary singular vector subsets, providing new algorithms with guarantees for matrix approximation problems and deepening the theoretical understanding of leverage-based methods.

Contribution

It introduces a generalized definition of leverage scores, connects them to principal angles, and develops approximation algorithms with provable guarantees for key matrix problems.

Findings

New bounds improve understanding of matrix approximations

Algorithms with provable guarantees for subset selection and CCA

Numerical experiments validate the proposed methods

Abstract

In problems involving matrix computations, the concept of leverage has found a large number of applications. In particular, leverage scores, which relate the columns of a matrix to the subspaces spanned by its leading singular vectors, are helpful in revealing column subsets to approximately factorize a matrix with quality guarantees. As such, they provide a solid foundation for a variety of machine-learning methods. In this paper we extend the definition of leverage scores to relate the columns of a matrix to arbitrary subsets of singular vectors. We establish a precise connection between column and singular-vector subsets, by relating the concepts of leverage scores and principal angles between subspaces. We employ this result to design approximation algorithms with provable guarantees for two well-known problems: generalized column subset selection and sparse canonical correlation analysis. We run numerical experiments to provide further insight on the proposed methods. The novel bounds we derive improve our understanding of fundamental concepts in matrix approximations. In addition, our insights may serve as building blocks for further contributions.