CS378 Computational Intelligence in Game Research

Spring 2012, JES A216A, TTh 3:30-5pm, Unique 53545

Joel Lehman
Office hours: Tu11-12,MW2-3 (GDC 2.212).


Minwoo Bae
Office hours: ENS 31NN, MWF8:50-9:50AM
Evaluation hours: ENS 31NN, TTh:5-6PM

Christopher Donahue
Office hours: ENS 31NN TuTh:1-2PM, F:2-3PM
Evaluation hours: ENS31NN, MW:10-11AM


The class reviews technologies that connect artificial intelligence and game programming. Two key tools are the NERO video game (www.nerogame.org), an award-winning game AI project that has been featured on Slashdot, Games Digest, AAAI, KXAN, and other venues; and OpenNERO (opennero.googlecode.com), an open source game platform for AI research and education featured in the most recent edition of Artificial Intelligence: A Modern Approach by Stuart Russell and Peter Norvig.

Students will become familiar with next-generation AI techniques that may play a central role in the next generation of video games and learn hands-on about working with modern game engines. In addition, teams of students will propose and work on original research projects aiming to improve the state of the art of AI in games.

There are no specific prerequisites, however you should be excited about working in some aspect of game technology and dedicated to making the team project a success.

Topics that the class covers include:

  • Artificial Neural Networks
  • Evolutionary Algorithms
  • Game Programming
  • Video Game Engines
  • Neuroevolution
  • Artificial Intelligence

The class will be 3 credit hours. 75% of the grade is based on homework (each graded pass/fail), and 25% on a written project proposal (graded by percentage).

Passing each homework assignment involves completeness and correctness. Proposals will be graded by descriptiveness, writing clarity, presentation, and and the described project's feasibility.

Turning in the proposal or a homework late will reduce the grade 15% for the first 24hrs, 40% for the second, 75% for the third, and 100% after that. "Extra credit" on a homework is 15% (unless otherwise indicated in the assignment).

Important resources:

NERO development wiki
This wiki (http://z.cs.utexas.edu/users/nn/nero/wiki/) lists information on homework, the class schedule, and sign-ups for evaluations with the mentors.

Class piazza group
The piazza group will be the main means of communication for the class, so it is important to sign up and check it regularly (http://piazza.com/utexas/spring2013/cs378/home).

Other useful links:
The UTCS Neural Networks Research Group
The NERO website

Notice: Students with disabilities may request appropriate academic accommodations from the Division of Diversity and Community Engagement, Services for Students with Disabilities, 512-471-6259, http://www.utexas.edu/diversity/ddce/ssd/