Convergence of Batch Asynchronous Stochastic Approximation With Applications to Reinforcement Learning

TL;DR

This paper studies the convergence properties of Block Asynchronous Stochastic Approximation (BASA), a generalization of stochastic approximation methods, with applications to reinforcement learning, providing conditions and bounds for convergence and rate.

Contribution

It introduces a unified framework for analyzing both synchronous and asynchronous stochastic approximation, including new convergence conditions and rate bounds for BASA.

Findings

Established convergence conditions for BASA.

Derived bounds on the rate of convergence.

Applied results to reinforcement learning fixed point problems.

Abstract

We begin by briefly surveying some results on the convergence of the Stochastic Gradient Descent (SGD) Method, proved in a companion paper by the present authors. These results are based on viewing SGD as a version of Stochastic Approximation (SA). Ever since its introduction in the classic paper of Robbins and Monro in 1951, SA has become a standard tool for finding a solution of an equation of the form $f(\theta) = 0$, when only noisy measurements of $f(\cdot)$ are available. In most situations, \textit{every component} of the putative solution $\theta_t$ is updated at each step $t$. In some applications in Reinforcement Learning (RL), \textit{only one component} of $\theta_t$ is updated at each $t$. This is known as \textbf{asynchronous} SA. In this paper, we study \textbf{Block Asynchronous SA (BASA)}, in which, at each step $t$, \textit{some but not necessarily all} components of $\theta_t$ are updated. The theory presented here embraces both conventional (synchronous) SA as well as asynchronous SA, and all in-between possibilities. We provide sufficient conditions for the convergence of BASA, and also prove bounds on the \textit{rate} of convergence of $\theta_t$ to the solution. For the case of conventional SGD, these results reduce to those proved in our companion paper. Then we apply these results to the problem of finding a fixed point of a map with only noisy measurements. This problem arises frequently in RL. We prove sufficient conditions for convergence as well as estimates for the rate of convergence.