arXiv (Cornell University) · 2020 · 45 citations · 24 references
ComputingComputer ScienceArtificial IntelligenceMachine LearningOptimizationContinuous OptimizationDeep LearningNeural NetworksPolicy LearningReinforcement LearningValue FunctionsBias MitigationDeep Reinforcement LearningOptimization AlgorithmsPolicy GradientsReinforcementValue Function Estimation
A widely-used actor-critic reinforcement learning algorithm for continuous control, Deep Deterministic Policy Gradients (DDPG), suffers from the overestimation problem, which can negatively affect the performance. Although the state-of-the-art Twin Delayed Deep Deterministic Policy Gradient (TD3) algorithm mitigates the overestimation issue, it can lead to a large underestimation bias. In this paper, we propose to use the Boltzmann softmax operator for value function estimation in continuous control. We first theoretically analyze the softmax operator in continuous action space. Then, we uncover an important property of the softmax operator in actor-critic algorithms, i.e., it helps to smooth the optimization landscape, which sheds new light on the benefits of the operator. We also design two new algorithms, Softmax Deep Deterministic Policy Gradients (SD2) and Softmax Deep Double Deterministic Policy Gradients (SD3), by building the softmax operator upon single and double estimators, which can effectively improve the overestimation and underestimation bias. We conduct extensive experiments on challenging continuous control tasks, and results show that SD3 outperforms state-of-the-art methods.
24
Human-level control through deep reinforcement learning
Volodymyr Mnih, Koray Kavukcuoglu, David Silver et al. · Nature · 2015 · 30K citations
Reinforcement Learning: An Introduction
IEEE Transactions on Neural Networks · 2005 · 25.8K citations
Mastering the game of Go with deep neural networks and tree search
David Silver, Aja Huang, Chris J. Maddison et al. · Nature · 2016 · 15.8K citations
Continuous control with deep reinforcement learning
Timothy Lillicrap, Jonathan J. Hunt, Alexander Pritzel et al. · arXiv (Cornell University) · 2015 · 5.4K citations · Full text
Computing, Computer Science, Physics +17