Xianyuan Zhan
TalkRL: The Reinforcement Learning Podcast
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.
00:00
Transcript
Play full episode
Remember Everything You Learn from Podcasts
Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
