图像索引>> Mountain_Car-using-DDQN>> 返回
项目作者： aksayushx

项目描述 ：
Deep RL agent for solving MountainCar-v0 environment.
高级语言： Python
项目地址: git://github.com/aksayushx/Mountain_Car-using-DDQN.git
创建时间： 2020-06-08T11:04:47Z
项目社区：https://github.com/aksayushx/Mountain_Car-using-DDQN
开源协议：
下载


Mountain Car using Deep RL

Environment Details

A car is on a one-dimensional track, positioned between two “mountains”. The goal is to drive up the mountain on the right; however, the car’s engine is not strong enough to scale the mountain in a single pass. Therefore, the only way to succeed is to drive back and forth to build up momentum.

Solved Requirements

MountainCar-v0 defines “solving” as getting average reward of -110.0 over 100 consecutive trials.

The target_dqn.h5 model gets an average score of -99.67 in 100 consecutive episodes ans thus solves the Mountain Car environment.

Performance Of Trained Model

After 750 episodes of training

Files

  1. target_dqn.h5 : A trained model for predicting Q values for state action pairs.
  2. train.py : Contains detailed implementation of the DDQN for solving Mountain Car.
  3. q_network.py : Contains the neural network architechture for predicting Q values.
  4. test.py : Runs the trained model on the environment for 100 episodes and returns the mean score.

Testing

  1. Epoch :88 Score :-104
  2. Epoch :89 Score :-131
  3. Epoch :90 Score :-104
  4. Epoch :91 Score :-104
  5. Epoch :92 Score :-88
  6. Epoch :93 Score :-130
  7. Epoch :94 Score :-103
  8. Epoch :95 Score :-85
  9. Epoch :96 Score :-104
  10. Epoch :97 Score :-104
  11. Epoch :98 Score :-102
  12. Epoch :99 Score :-129
  14. Average Reward  in 100 games : -100.58