fyng / RL

Usage

Install micromamba or mamba as the package manager. To install micromamba, refer to the installation guide

To install the classifier:

1. Clone the repo

git clone https://github.com/fyng/RL.git
cd RL

2. Create virtual environment

micromamba env create -f environment.yml
micromamba activate RL

3. Create directory plots for plots

4. Run model

• Question 1: Run Maze/value_iter.py to generate Q and V. Run Maze/value_iter_demo.py to visualize the optimal path through the maze.
• Question 2: Run Maze/qlearning_maze.py to generate plots comparing Q-learning against Q* found through value iteration in question 1.
• Question 3: Run Gym/reinforce.py for the REINFORCE algorithm, or Gym\qlearning.py for Q-learning. Uncomment the appropriate lines for Acrobat-V1 and Mountaincar-V0.

Acknowledgement

ECE 5242 - Intelligent Autonomous Systems taught by Prof Daniel Lee & Travers Rhodes