A Max-Min Entropy Framework for Reinforcement Learning

In this paper, we propose a max-min entropy framework for reinforcement learning (RL) to overcome the limitation of the soft actor-critic (SAC) algorithm implementing the maximum entropy RL in model-free sample-based learning. Whereas the maximum entropy RL guides learning for policies to reach stat...

Full description

Saved in:
Bibliographic Details
Published inarXiv.org
Main Authors Han, Seungyul, Sung, Youngchul
Format Paper
LanguageEnglish
Published Ithaca Cornell University Library, arXiv.org 20.12.2021
Subjects
Algorithms
Entropy
Learning
Markov processes
Maximum entropy
Online AccessGet full text

Cover

More Information
Summary:In this paper, we propose a max-min entropy framework for reinforcement learning (RL) to overcome the limitation of the soft actor-critic (SAC) algorithm implementing the maximum entropy RL in model-free sample-based learning. Whereas the maximum entropy RL guides learning for policies to reach states with high entropy in the future, the proposed max-min entropy framework aims to learn to visit states with low entropy and maximize the entropy of these low-entropy states to promote better exploration. For general Markov decision processes (MDPs), an efficient algorithm is constructed under the proposed max-min entropy framework based on disentanglement of exploration and exploitation. Numerical results show that the proposed algorithm yields drastic performance improvement over the current state-of-the-art RL algorithms.
ISSN:2331-8422