CS-0277: Artificial Neural Networks Control for Game Playing Agents

This course will combine work on two important subfields in artificial intelligence: artificial neural networks, and artificial agents. Artificial Neural Networks are computational devices loosely based on the brain. Agent-based systems use a collection of information to solve a complex task, while possibly providing for planning, communication, error recovery, and learning. In this course we will combine these two techniques and apply them to virtual simulations of games such as Capture the Flag, Robocup, RoboRescue, Quidditch, and others. Emphasis will be placed on the design and implementation of course projects. Although several neural network and virtual environment programs will be made available to students, developing individual projects will require making modifications to (and/or interfaces for) these programs. Prerequisite: students are expected to have at least a semester worth of programming experience before this course.