CC+ CS 4701 Ningchuan Wan William Moseson Swinburn Miranda
Problem Statement and Motivation • To develop an artificial Chinese Checkers(modified) player • Motivated by the experience gained and bragging rights.
Input/Output Specification • Our program is initially given the board state. • After each round, the server gives, as input, the last move performed. • Our program requests a desired move to the server as output.
Approach • Search with Minimax determined by heuristic • Edge weights are determined by heuristics • Distance of the piece from the goal • Potential travel distance • Avoiding stragglers • Learning based on • Training data • Generalization of states
Where is the AI? • MiniMax Algorithm • Heuristics and strategies • Learning from external players and opponent
SystemArchitecture Decision module I/O module Communication Server Knowledge base Log Learning Engine