使用Gym库+keras实现强化学习算法

冷不防 2023-06-21 07:17 50阅读 0赞

【时间】2019.12.11

【题目】使用Gym库+keras实现强化学习算法

### 一、Gym库介绍与使用 ###

[gym 介绍][gym]

### 二、例子 ###

[Keras深度强化学习-- Policy Network与DQN实现][Keras_-- Policy Network_DQN]

相关资源：

**Paper**：  
Policy Gradient:[Policy gradient methods for reinforcement learning with function approximation][]  
DQN:[ Playing Atari with Deep Reinforcement Learning][Playing Atari with Deep Reinforcement Learning]  
NatureDQN:[Human-level control through deep reinforcement learning][]

> **Github**：[https://github.com/xiaochus/Deep-Reinforcement-Learning-Practice][https_github.com_xiaochus_Deep-Reinforcement-Learning-Practice]
> 
> 这是一个用keras实现的深度强化学习相关算法的资源code

## Algorithm ##

<table> 
 <thead> 
  <tr> 
   <th>#</th> 
   <th>Name</th> 
   <th>Paper</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td>1</td> 
   <td><a href="https://github.com/xiaochus/Deep-Reinforcement-Learning-Practice/blob/master/DQN.py" rel="nofollow">DQN</a></td> 
   <td><a href="http://cn.arxiv.org/abs/1312.5602" rel="nofollow">Playing Atari with Deep Reinforcement Learning</a></td> 
  </tr> 
  <tr> 
   <td>2</td> 
   <td><a href="https://github.com/xiaochus/Deep-Reinforcement-Learning-Practice/blob/master/NatureDQN.py" rel="nofollow">Nature DQN</a></td> 
   <td><a href="https://www.nature.com/articles/nature14236" rel="nofollow">Human-level control through deep reinforcement learning</a></td> 
  </tr> 
  <tr> 
   <td>3</td> 
   <td><a href="https://github.com/xiaochus/Deep-Reinforcement-Learning-Practice/blob/master/DoubleDQN.py" rel="nofollow">Double DQN</a></td> 
   <td><a href="http://cn.arxiv.org/abs/1509.06461v1" rel="nofollow">Deep Reinforcement Learning with Double Q-learning</a></td> 
  </tr> 
  <tr> 
   <td>4</td> 
   <td><a href="https://github.com/xiaochus/Deep-Reinforcement-Learning-Practice/blob/master/DuelingDQN.py" rel="nofollow">Dueling DQN</a></td> 
   <td><a href="https://arxiv.org/pdf/1511.06581.pdf" rel="nofollow">Dueling Network Architectures for Deep Reinforcement Learning</a></td> 
  </tr> 
  <tr> 
   <td>5</td> 
   <td><a href="https://github.com/xiaochus/Deep-Reinforcement-Learning-Practice/blob/master/AC_sparse.py" rel="nofollow">Actor-Critic</a></td> 
   <td><a href="https://papers.nips.cc/paper/1786-actor-critic-algorithms.pdf" rel="nofollow">Actor-Critic Algorithms</a></td> 
  </tr> 
  <tr> 
   <td>6</td> 
   <td><a href="https://github.com/xiaochus/Deep-Reinforcement-Learning-Practice/blob/master/PolicyNetwork.py" rel="nofollow">Policy Network</a></td> 
   <td><a href="https://www.researchgate.net/publication/2503757_Policy_Gradient_Methods_for_Reinforcement_Learning_with_Function_Approximation" rel="nofollow">Policy gradient methods for reinforcement learning with function approximation</a></td> 
  </tr> 
  <tr> 
   <td>7</td> 
   <td><a href="https://github.com/xiaochus/Deep-Reinforcement-Learning-Practice/blob/master/DDPG.py" rel="nofollow">DDPG</a></td> 
   <td><a href="https://arxiv.org/abs/1509.02971" rel="nofollow">Continuous Control with Deep Reinforcement Learning</a></td> 
  </tr> 
  <tr> 
   <td>8</td> 
   <td><a href="https://github.com/xiaochus/Deep-Reinforcement-Learning-Practice/blob/master/PPO_TF.py" rel="nofollow">PPO_TF</a></td> 
   <td><a href="https://arxiv.org/abs/1707.06347" rel="nofollow">Proximal Policy Optimization Algorithms</a></td> 
  </tr> 
  <tr> 
   <td>9</td> 
   <td><a href="https://github.com/xiaochus/Deep-Reinforcement-Learning-Practice/blob/master/A3C_sparse.py" rel="nofollow">A3C</a></td> 
   <td><a href="https://arxiv.org/pdf/1602.01783.pdf" rel="nofollow">Asynchronous Methods for Deep Reinforcement Learning</a></td> 
  </tr> 
 </tbody> 
</table>

[gym]: https://www.jianshu.com/p/e136f4fba20c
[Keras_-- Policy Network_DQN]: https://www.jianshu.com/p/e037d42ab6b1
[Policy gradient methods for reinforcement learning with function approximation]: https://www.researchgate.net/publication/2503757_Policy_Gradient_Methods_for_Reinforcement_Learning_with_Function_Approximation
[Playing Atari with Deep Reinforcement Learning]: http://cn.arxiv.org/abs/1312.5602
[Human-level control through deep reinforcement learning]: https://www.nature.com/articles/nature14236
[https_github.com_xiaochus_Deep-Reinforcement-Learning-Practice]: https://github.com/xiaochus/Deep-Reinforcement-Learning-Practice