An Analysis of Reinforcement Learning with Function Approximation
Organizations: Carnegie Mellon University, Pittsburgh, PA 15213, USA · Coordinated Science Lab, Urbana, IL 61801, USA · Institute for Systems and Robotics, 1049-001 Lisboa, Portugal
Abstract
We address the problem of computing the optimal Q-function in Markov decision problems with infinite state-space. We analyze the convergence properties of several variations of Q-learning when combined with function approximation, extending the analysis of TD-learning in Tsitsiklis and Van Roy (1996) to stochastic control settings. We identify conditions under which such approximate methods converge with probability 1. We conclude with a brief discussion on the general applicability of our results and compare them with several related works.
Q-Learning
Q-learning estimates long-term action values from experience so agents can improve decisions without a learned transition model.
See all 114 papers