oyeshbase / multiagentpacman Goto Github PK
View Code? Open in Web Editor NEWThis project forked from jornason/multiagentpacman
This is the second project of the Spring 2014 CS 188 (Introduction to Artificial Intelligence) class at UC Berkeley. It implements a pacman who can play against adversarial agents with dynamic tree look up (both Minimax and Expectimax). it optimizes tree lookup with AlphaBetaPruning and finally contains a master strategy in the form of the betterEvaluationFunction (in multiAgents.py) for evaluating each state in the tree.