AI-SARAH: Adaptive and Implicit Stochastic Recursive Gradient Methods

Zheng Shi, Nicolas Loizou, Peter Richtárik, Martin Takac

Keywords: practical variant of SARAH, adaptive step-size, tune-free algorithm, implicit approach, convex optimization in machine learning

Abstract: We present AI-SARAH, a practical variant of SARAH. As a variant of SARAH, this algorithm employs the stochastic recursive gradient yet adjusts step-size based on local geometry. AI-SARAH implicitly computes step-size and efficiently estimates local Lipschitz smoothness of stochastic functions. It is fully adaptive, tune-free, straightforward to implement, and computationally efficient. We provide technical insight and intuitive illustrations on its design and convergence. We conduct extensive empirical analysis and demonstrate its strong performance compared with its classical counterparts and other state-of-the-art first-order methods in solving convex machine learning problems.

One-sentence Summary: A tune-free & fully adaptive algorithm and a practical variant of SARAH.