Lunar Lander Deep Reinforcement Learning

Train a deep reinforcement learning model to solve the Open AI Lunar Lander problem.

Environment

OpenAI Gym library Lunar Lander.
Code is mainly based on the following repos:

• Reinforcement-Learning-Lunar_Lander
• TensorflowKeras-ReinforcementLearning

Dependencies and Installation

Step 1:

• Microsoft C++ Build Tools: → "MSVCv142 - VS 2019 C++ x64/x86 build tools" or greater
• pip install gym
• pip install gym[box2d]
• pip install tensorflow==2.12.0
• pip install keras==2.12.0
• pip install keras-rl2

Step 2 (for DQNAgent Model):
Since the keras rl library (keras-rl2) is outdated and not updated anymore, some changed had to be made to the existing library files. In ...site-packages\rl replace the callbacks.py and core.py file with the files in this repo. Same for dqn.py in ...site-packages\rl\agents.

How to use

Code provides two classes/models for training the Open AI Lunar Lander: DQNAgent (from keras rl library) and custom DQN model (based on keras library). Both models use a greedy policy (more information at Reinforcement-Learning-Lunar_Lander). In the main file, the model as well as other options like rendering and testing can be selected.