Date of Original Version

2014

Type

Conference Proceeding

Journal Title

Journal of Machine Learning Research : Workshop and Conference Proceedings

Volume

32

First Page

1827

Last Page

1835

Rights Management

Copyright 2014 by the author(s).

Abstract or Description

In this paper, we consider the problem of finding a linear (binary) classifier or providing a near-infeasibility certificate if there is none. We bring a new perspective to addressing these two problems simultaneously in a single efficient process, by investigating a related Bilinear Saddle Point Problem (BSPP). More specifically, we show that a BSPP-based approach provides either a linear classifier or an ϵ-infeasibility certificate. We show that the accelerated primal-dual algorithm, Mirror Prox, can be used for this purpose and achieves the best known convergence rate of O(√logn/ρ(A)) (O(√logn/ϵ)), which is almost independent of the problem size, n. Our framework also solves kernelized and conic versions of the problem, with the same rate of convergence. We support our theoretical findings with an empirical study on synthetic and real data, highlighting the efficiency and numerical stability of our algorithms, especially on large-scale instances.

Share

COinS
 

Published In

Journal of Machine Learning Research : Workshop and Conference Proceedings, 32, 1827-1835.