随机过程

这里分享一下我制作的课件(全英文,30课时,打开密码请发邮件咨询)。主要介绍马尔科夫决策过程(Markov Decision Process)以及部分强化学习的知识,具体内容包括:

  1. Introduction
  • Course summary
  • Problems
  • Concepts
  • MDP elements
  • Agent-environment interface
  1. Finite-horizon MDP
  • Model formulation
  • Finite-horizon MDP
  • Solve four examples using dynamic programming
  • Monotone policies
  1. Infinite-horizon MDP
  • Introduction
  • Discounted MDP
  • Expected total-reward criterion
  1. ADP and RL
  • Introduction
  • Monte Carlo methods
  • Temporal Difference learning
  • Frontiers