Artificial Intelligence: foundations of computational agents

Interactive Demos

David Poole

These applets are to demonstrate some of the algorithms in Artificial Intelligence: foundations of computational agents . They are all copyright by David Poole and released under the GPL.

Hidden Markov Models

• Bayesian Localization demo, (See also Sebastian Thrun's Monte Carlo Localization videos)

Bayesian Learning

• Beta distribution

Decision-theoretic Planning

• Value iteration.

Multi-Agent Systems

• Policy Improvement.

Reinforcement Learning

• Gridworld Q-learning.
• Gridworld SARSA-lambda.
• A Tiny Game with 6 states and 4 actions, with Q-learning and SARSA

• A Simple Game with a rather simplistic rule-based controller
• Q-learning controller
• Linear function controller -- SARSA with linear function approximation
• Adversary controller (Q-learning, but where an adversary chooses the prize location).

David Poole - 2007-03-26