Demonstrated value iteration and Q-learning onto agents in a grid world, as well as simulated robot crawler and Pac-Man
This project was written in Python 3.6
To run gridworld and control the agent manually, run command: python gridworld.py -m
To see a full list of options for running gridworld, run command: python gridworld.py -h
To see a gridworld agent run with value iteration, run command: python autograder.py -q q1
To play around and visualize the differences in policies depending on inputted MDP values, run command: python gridworld.py -g DiscountGrid -a value --discount [YOUR_DISCOUNT] --noise [YOUR_NOISE] --livingReward [YOUR_LIVING_REWARD]
To manually control and watch Q-learning agent, run command: python gridworld.py -a q -k 5 -m
To observe Q-learning agent with Epsilon Greedy action selection, run commands: python gridworld.py -a q -k 100 --noise 0.0 -e 0.1 python gridworld.py -a q -k 100 --noise 0.0 -e 0.9
To observe the Q-learning crawler robot, run command: python crawler.py
To make Pac-Man use Q-learning in order to complete the game, run command: python autograder.py -q q5
If you want to see Pac-Man's Q-learning visually, run command: python pacman.py -p PacmanQAgent -n 10 -l smallGrid -a numTraining=10
If you want to see Pac-Man use approximate Q-learning, run commands: python pacman.py -p ApproximateQAgent -a extractor=SimpleExtractor -x 50 -n 60 -l mediumGrid python pacman.py -p ApproximateQAgent -a extractor=SimpleExtractor -x 50 -n 60 -l mediumClassic