Welcome to lunarlander_gym’s documentation!¶
Summary¶
This project is implementation of multiple AI agents based on different Reinforcement Learning methods to OpenAI Gymnasium Lunar-Lander environment which is classic rocket landing trajectory optimization problem.
Free software: MIT license
Home Page: https://github.com/Ehsan2754/lunarlander_gym.
Evaluation¶
RandomAgent |
Gradient Policy Agent |
Q-Learning Agent |
A ctor-critic Agent |
|
|---|---|---|---|---|
Training episodes |
0 |
10,000 |
3000 |
3000 |
Reward |
-70.46 |
49.07 |
198.51 |
284.86 |