Click here for our new site

Machine Learning

Reid M Knight

For many problems in AI, and games in particular, it is beneficial to allow an agent (general name for a program that runs with an artificial intelligence) some form of predictive power, and the ability to learn a certain amount of information. While the field of machine learning is rather broad scoped, this talk will focus primarily on two different approaches - Decision Trees and Reinforcement Learning.

The talk assumes that you have a brief understanding of artificial intelligence concepts, as well as understanding of a Tree structure (whether or not you understand how to code one in software).

Fork me on GitHub