On Convergence of Average-Reward Q-Learning in Weakly Communicating Markov Decision Processes

TL;DR

This paper extends the convergence analysis of RVI Q-learning algorithms to weakly communicating MDPs, broadening theoretical understanding and practical applicability in average-reward reinforcement learning.

Contribution

It generalizes convergence proofs from unichain to weakly communicating MDPs and characterizes the convergence sets for RVI Q-learning algorithms.

Findings

RVI Q-learning converges almost surely in weakly communicating MDPs.

The convergence set is compact, connected, and solutions to the average-reward optimality equation.

Hierarchical RVI algorithms also converge under weakly communicating assumptions.

Abstract

This paper analyzes reinforcement learning (RL) algorithms for Markov decision processes (MDPs) under the average-reward criterion. We focus on Q-learning algorithms based on relative value iteration (RVI), which are model-free stochastic analogues of the classical RVI method for average-reward MDPs. These algorithms have low per-iteration complexity, making them well-suited for large state space problems. We extend the almost-sure convergence analysis of RVI Q-learning algorithms developed by Abounadi, Bertsekas, and Borkar (2001) from unichain to weakly communicating MDPs. This extension is important both practically and theoretically: weakly communicating MDPs cover a much broader range of applications compared to unichain MDPs, and their optimality equations have a richer solution structure (with multiple degrees of freedom), introducing additional complexity in proving algorithmic convergence. We also characterize the sets to which RVI Q-learning algorithms converge, showing that they are compact, connected, potentially nonconvex, and comprised of solutions to the average-reward optimality equation, with exactly one less degree of freedom than the general solution set of this equation. Furthermore, we extend our analysis to two RVI-based hierarchical average-reward RL algorithms using the options framework, proving their almost-sure convergence and characterizing their sets of convergence under the assumption that the underlying semi-Markov decision process is weakly communicating.