This project, by Anne M. Tumlin of Vanderbilt University, compares Reinforcement Learning (RL) and Linear Control Methods, focusing on the Linear Quadratic Regulator (LQR) and Q-Learning approach for solving the inverted pendulum problem.
- Python
- MATLAB
- OpenAI Gym (for RL simulations)
Run the Python notebook for the RL methods and MATLAB for LQR.
Overview of RL techniques (Q-Learning, Deep Q-Learning, Double Q-Learning) and LQR. See the project report for more details.
Compares effectiveness, reliability, and adaptability of RL and LQR in maintaining pendulum balance and safety. See the project report for more details.
A report detailing the findings and methods of the project can be found in the repository under CS6376_FinalProjectPaper.pdf.
A video walk-through demonstrating the code and results can be found at this link: Code Demonstration
Anne M. Tumlin, Vanderbilt University
Contact the author for inquiries or refer to the project document for detailed information.