Implicit Distributional Reinforcement Learning

TL;DR

This paper introduces Implicit Distributional Actor-Critic (IDAC), a reinforcement learning method that models complex return distributions and policies using implicit and semi-implicit distributions, leading to improved sample efficiency and performance in continuous action tasks.

Contribution

The paper proposes a novel implicit distributional approach with deep generator networks and semi-implicit policies, enhancing modeling flexibility and performance in policy-gradient reinforcement learning.

Findings

IDAC outperforms state-of-the-art algorithms on OpenAI Gym benchmarks.

The implicit distributional approach captures complex return and policy properties.

IDAC demonstrates improved sample efficiency in continuous control tasks.

Abstract

To improve the sample efficiency of policy-gradient based reinforcement learning algorithms, we propose implicit distributional actor-critic (IDAC) that consists of a distributional critic, built on two deep generator networks (DGNs), and a semi-implicit actor (SIA), powered by a flexible policy distribution. We adopt a distributional perspective on the discounted cumulative return and model it with a state-action-dependent implicit distribution, which is approximated by the DGNs that take state-action pairs and random noises as their input. Moreover, we use the SIA to provide a semi-implicit policy distribution, which mixes the policy parameters with a reparameterizable distribution that is not constrained by an analytic density function. In this way, the policy's marginal distribution is implicit, providing the potential to model complex properties such as covariance structure and skewness, but its parameter and entropy can still be estimated. We incorporate these features with an off-policy algorithm framework to solve problems with continuous action space and compare IDAC with state-of-the-art algorithms on representative OpenAI Gym environments. We observe that IDAC outperforms these baselines in most tasks. Python code is provided.