Xianyuan Zhan
TalkRL: The Reinforcement Learning Podcast
00:00
Understanding the MORE Algorithm: A Model-Based Approach to Offline Reinforcement Learning
This chapter examines the MORE algorithm, a model-based offline reinforcement learning strategy tailored for high-dimensional state and action spaces in thermal power plants. It discusses the development of a data-driven simulator, advanced data processing techniques, and the integration of safety constraints in policy optimization.
Transcript
Play full episode
