策略引擎>> rl-code>> 返回
项目作者： KibaAmor

项目描述 ：
Reinforcement Learning Code Snippet
高级语言： Python
项目地址: git://github.com/KibaAmor/rl-code.git
创建时间： 2020-08-10T07:44:33Z
项目社区：https://github.com/KibaAmor/rl-code
开源协议：MIT License
下载


Reinforcement Learning Code Snippet

Understand the basic concepts in reinforcement learning

HowToRun

  1. install requirements

    1.  pip install -r requirements.txt

  2. run code

    1.  cd value-based/tabular
    3.  # test sarsa code
    4.  python -m sarsa
    5.  # if running in terminal without display, use blow command
    6.  xvfb-run -a -s "-screen 0 1400x900x24" python -m raw_dqn

1. Value Based Learning

1.1 Tabular Based

  1. Sarsa
  2. Q-Learning

1.2 Nerual Network Based

  1. DQN
  2. Double DQN

2. Policy Based Learning

3. Actor-Critic