强化学习

In order to realize autonomous systems that can learn to make correct decisions, reinforcement learning is a powerful paradigm that is related to many tasks. This course will provide a solid introduction to the field of reinforcement learning, and the students will learn the core challenges and methods, including Markov decision process, dynamic programming, mode-free prediction and control, value function approximation, policy gradient, learning and planning, deep reinforcement learning and multi-agent reinforcement learning. The goal is to provide the basic ideas and methods in reinforcement learning, and the students will be able to pursue advanced study and research in the field if desired.