This project demonstrates various Artificial Intelligence techniques such as searching, adversarial behaviour, deep reinforcement learning, neural networks etc. to train a Pacman agent allowing it to maximize its expected utility.
- Project 0 - Introduction/Tutorial to Python
- Project 1 - search - DFS, BFS, A*, etc.
- Project 2 - multiagent - minimax, alpha-beta pruning, expectimax, etc.
- Project 3 - reinforcement learning - MDP, value iteration, q-learning, epsilon-greedy, approximate q-learning, etc.
- Project 4 - tracking - HMM, particle filtering, bayes' nets, deep reinforcement learning, etc.
- Pacman Competition - The goal is to eat the other pacman before it eats you.
For more details please check the specification present under "specification" folder in each project directory.
Programming Language - Python 2.7