If this code helped you save time, help me out by clicking the star at the top!
Project from AeroE 504: Decision Making under Uncertainty. Course taught by Professor Peng Wei. There are 3 different datasets we were given to work with:
- Small.csv - Gridworld environment
- Medium.csv - MountainCar Continuous environment
- Large.csv - Unknown environment
For the different envrionments, I implement 2 different reinforcement learning algorithms to solve the problems.
- Small.csv was solved using maximum likelihood estimation combined with value iteration. It converges to the optimal policy.
- Medium.csv and Large.csv was solved using a temporal difference Q-learning and creating episodes to split the data. It converged to a policy that performed very well in the MountainCar Environment.