Reinforcement learning investigation of SARSA, Q-Learning, and Monte Carlo Tree Search for Minesweeper with Markov Decision Processes
WINTER 2023
A project originally used for the CS 238: Decision Making Under Uncertainty final project. The project investigated reinforcement learning and policy search on the game Minesweeper. Experiments included investigations into SARA, Q-Learning, and Monte Carlo Tree Search algorithms and implementations. Pair-Programmed code can be found at https://github.com/benjaminhwittenbrink/cs238project
Topics covered and tools used:
- Reinforment Learning
- Large State Space Representation
- Markov Decision Processes
- Q-Search
- SARSA
- Monte Carlo Tree Search
- Exploitation vs Exploration