Gradual Transition from Bellman Optimality Operator to Bellman Operator in Online Reinforcement Learning

TL;DR

This paper introduces an annealing method that gradually shifts from the Bellman optimality operator to the Bellman operator in actor-critic RL algorithms, improving learning speed and reducing bias in continuous action spaces.

Contribution

The study proposes a novel annealing approach to transition between Bellman operators, enhancing sample efficiency and robustness in continuous action reinforcement learning.

Findings

Modeling optimal values accelerates learning.

The annealing method reduces overestimation bias.

The approach outperforms existing methods in various tasks.

Abstract

For continuous action spaces, actor-critic methods are widely used in online reinforcement learning (RL). However, unlike RL algorithms for discrete actions, which generally model the optimal value function using the Bellman optimality operator, RL algorithms for continuous actions typically model Q-values for the current policy using the Bellman operator. These algorithms for continuous actions rely exclusively on policy updates for improvement, which often results in low sample efficiency. This study examines the effectiveness of incorporating the Bellman optimality operator into actor-critic frameworks. Experiments in a simple environment show that modeling optimal values accelerates learning but leads to overestimation bias. To address this, we propose an annealing approach that gradually transitions from the Bellman optimality operator to the Bellman operator, thereby accelerating learning while mitigating bias. Our method, combined with TD3 and SAC, significantly outperforms existing approaches across various locomotion and manipulation tasks, demonstrating improved performance and robustness to hyperparameters related to optimality. The code for this study is available at https://github.com/motokiomura/annealed-q-learning.