Tencent AI Lab 官网
Geometric Descent Method for Convex Composite Minimization
Abstract View the Paper

In this paper, we extend the geometric descent method recently proposed by Bubeck, Lee and Singh to tackle nonsmooth and strongly convex composite problems. We prove that our proposed algorithm, dubbed geometric proximal gradient method (GeoPG), converges with a linear rate (1-1⁄√κ) and thus achieves the optimal rate among first-order methods, where κ is the condition number of the problem. Numerical results on linear regression and logistic regression with elastic net regularization show that GeoPG compares favorably with Nesterov’s accelerated proximal gradient method, especially when the problem is ill-conditioned.

Venue
2017 NIPS
Publication Time
Dec 2017
Authors
Shixiang Chen, Shiqian Ma, Wei Liu