Is Value Functions Estimation with Classification Plug-and-Play for Offline Reinforcement Learning?
Published in Automated Reinforcement Learning Workshop, 2024
This paper examines whether classification-based value function estimation can be used as a plug-and-play component in offline reinforcement learning pipelines.
Recommended citation: Denis Tarasov, Kirill Brilliantov, and Dmitrii Kharlapenko. "Is Value Functions Estimation with Classification Plug-and-Play for Offline Reinforcement Learning?" Automated Reinforcement Learning Workshop, 2024.
Download Paper