Parallel distributed processing: explorations in the microstructure of cognition, vol. 1: foundations
Temporal difference learning and TD-Gammon
Communications of the ACM
Artificial Neural Networks: A Tutorial
Computer - Special issue: neural computing: companion issue to Spring 1996 IEEE Computational Science & Engineering
Hi-index | 0.00 |
Determining a strong strategy for a game is a difficult task. What are good heuristics to use? What kind of states should be considered valuable or detrimental? With practice and analysis, a human can acquire this knowledge; is the same possible for a computer? Certainly a computer can be programmed with seemingly effective heuristics, but can a computer learn on its own what moves lead to victory and what others lead to defeat? The goal of this project is to take a computer with zero strategic knowledge of a game (for this project, Othello) and have it become on its own as great a player as it can without any external force guiding it. The approach taken to achieve this objective is to use an artificial neural network and train it with reinforcement learning techniques.