Deep Q-Network (DQN)

Installation

Prerequisites

To run this project, you will need to have anaconda or mini conda already installed: create a virtual environment with the following. To begin open your anaconda terminal and clone this repository

git clone https://github.com/sokistar24/Deep_Reinforcement_learning

conda create --name intro_to_rl python=3.9 jupyterlab numpy matplotlib

Activate the environment uisng

conda activate intro_to_rl

pip install swig

pip install gymnasium[box2d]

To start the JupyterLab server, run the following command in the terminal or command prompt:

jupyter-lab

Instructions

In this exercise, you will implement Deep Q-Learning to solve OpenAI Gym's LunarLander environment.

navigate to the Deep_Q_lunar_landar.ipynb/ and run the notebook

try to change the parameters in the notebook, to see if you can get the agent to train faster! You may also like to implement prioritized experience replay, or use it as a starting point to implement a Double DQN or Dueling DQN!

Results

Resources

• Human-Level Control through Deep Reinforcement Learning
• Deep Reinforcement Learning with Double Q-Learning
• Dueling Network Architectures for Deep Reinforcement Learning
• Prioritized Experience Replay