AI algorithms for Pacman

Intro

The Pacman Projects by the University of California, Berkeley.

Start a game with the command and move the agents using ASWD keyboard buttons or arrow keys:

$ python pacman.py

You can see the list of all options and their default values via:

$ python pacman.py -h

Multi-Agent algorithms

• ReflexAgent: an agent that considers food locations, ghost locations and score to perform well.

$ python pacman.py -p ReflexAgent -l originalClassic -n 1 -ghost DirectionalGhost -z 0.8 -k 1

• MinimaxAgent: an adversarial search agent implementing minimax algorithm

$ python pacman.py -p MinimaxAgent -l minimaxClassic -a depth=4

• AlphaBetaAgent: an adversarial search agent implementing minimax algorithm with alpha-beta pruning to more efficiently explore the minimax tree.

$ python pacman.py -p AlphaBetaAgent -l openClassic -a depth=2

• Expectimax: an adversarial search agent implementing expectimax algorithm

$ python pacman.py -l mediumClassic -p ExpectimaxAgent -a depth=2

Search algorithms

• DeepSearch: a deep search algorithm to find the best possible path given an evaluation function, it si faster than minimax but doesn't keep into considerations ghosts

$ python pacman.py -l trickyClassic -p DeepSearchAgent -a depth=6 evalFn=evaluationFunction