项目作者: angerhang

项目描述 :
RL-cheat-sheet
高级语言: TeX
项目地址: git://github.com/angerhang/reinforcement-learning-cheat-sheet.git
创建时间: 2019-02-27T09:54:32Z
项目社区:https://github.com/angerhang/reinforcement-learning-cheat-sheet

开源协议:Apache License 2.0

下载


reinforcement-learning-cheat-sheet

As part of my master thesis in reinforcement learning, I am gradually putting most of the basic notations all the way to the popular state-of-art methods in one place. Any suggestion or contribution is welcome :

Download

To download the complete cheat sheet in the pdf format, please click here

To Do

  • Basic notations
  • Problem setup
  • Policy gradient methods
  • Policy gradient vs Value-based
  • TD
  • SARSA
  • On-policy vs off-policy
  • Model-based vs Model-free
  • Dynamic programming
  • DQN
  • Double DQN
  • Priority replay
  • A3C
  • Distributed DQN
  • Noisy DQN
  • Recurrent DQN
  • Dueling Arch
  • Experience replay

References

This cheat sheet is written based on: