Estimation Error Correction in Deep Reinforcement Learning for Deterministic Actor-Critic Methods

Abstract

In value-based deep reinforcement learning methods, approximation of value functions induces overestimation bias and leads to suboptimal policies. We show that in deep actor-critic methods that aim to overcome the overestimation bias, if the reinforcement signals received by the agent have a high variance, a significant underestimation bias arises. To minimize the underestimation, we introduce a parameter-free, novel deep Q-learning variant. Our Q-value update rule combines the notions behind Clipped Double Q-learning and Maxmin Q-learning by computing the critic objective through the nested combination of maximum and minimum operators to bound the approximate value estimates. We evaluate our modification on the suite of several OpenAI Gym continuous control tasks, improving the state-of-the-art in every environment tested.

References in corpus (11)

• Continuous control with deep reinforcement learning
• Playing Atari with Deep Reinforcement Learning
• Addressing Function Approximation Error in Actor-Critic Methods
• OpenAI Gym
• IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures
• Off-Policy Deep Reinforcement Learning without Exploration
• Safe and Efficient Off-Policy Reinforcement Learning
• Learning to Play in a Day: Faster Deep Reinforcement Learning by Optimality Tightening
• Mean-Variance Optimization in Markov Decision Processes
• WD3: Taming the Estimation Bias in Deep Reinforcement Learning
• Learning Value Functions in Deep Policy Gradients using Residual Variance