Discrete Action On-Policy Learning with Action-Value Critic

Abstract

Reinforcement learning (RL) in discrete action space is ubiquitous in real-world applications, but its complexity grows exponentially with the action-space dimension, making it challenging to apply existing on-policy gradient based deep RL algorithms efficiently. To effectively operate in multidimensional discrete action spaces, we construct a critic to estimate action-value functions, apply it on correlated actions, and combine these critic estimated action values to control the variance of gradient estimation. We follow rigorous statistical analysis to design how to generate and combine these correlated actions, and how to sparsify the gradients by shutting down the contributions from certain dimensions. These efforts result in a new discrete action on-policy RL algorithm that empirically outperforms related on-policy algorithms relying on variance control techniques. We demonstrate these properties on OpenAI Gym benchmark tasks, and illustrate how discretizing the action space could benefit the exploration phase and hence facilitate convergence to a better local optimal solution thanks to the flexibility of discrete policy.

References in corpus (17)

• Auto-Encoding Variational Bayes
• Continuous control with deep reinforcement learning
• Playing Atari with Deep Reinforcement Learning
• Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor
• Trust Region Policy Optimization
• Addressing Function Approximation Error in Actor-Critic Methods
• High-Dimensional Continuous Control Using Generalized Advantage Estimation
• Reinforcement Learning with Deep Energy-Based Policies
• Sample Efficient Actor-Critic with Experience Replay
• REBAR: Low-variance, unbiased gradient estimates for discrete latent variable models
• Q-Prop: Sample-Efficient Policy Gradient with An Off-Policy Critic
• Backpropagation through the Void: Optimizing control variates for black-box gradient estimation
• Variance Reduction for Policy Gradient with Action-Dependent Factorized Baselines
• Off-Policy Actor-Critic
• Action-depedent Control Variates for Policy Optimization via Stein's Identity
• Discretizing Continuous Action Space for On-Policy Optimization
• ARSM: Augment-REINFORCE-Swap-Merge Estimator for Gradient Backpropagation Through Categorical Variables