Instructor: David Roberts
Time: TTh 12:50--2:05
As hardware technologies mature and players' expectations of computer games and entertainment experiences increase, the challenges for designers continue to grow as well. Modern computer games ship only after the efforts of large teams of artists, coders, and designers are coordinated to produce a coherent product. These teams are needed to produce artifacts of increasing complexity and scope that have come to be expected in computer games. Game designers and academics alike have been increasingly turning toward artificial intelligence techniques as tools to both create more engaging/interesting/complex game experiences and to enable the teams of programmers to produce these experiences with less effort.
In this course we will examine AI and ML algorithms for computer games as well as techniques for effectively evaluating these algorithms using players rather than simulations. We will look both at how these algorithms can be used to engineer intelligence into games and how these methods can reduce the effort required to design and implement games. Topics involving user studies and human computer interaction techniques will also be discussed, including study design, evaluation metrics, and basic statistics for data analysis. Although there are games in which computers play head-to-head, in this course we will focus on games as a playable artifact for humans. In doing so, we will study evaluation methodologies for the AI and ML techniques discussed in the course from a user's (either player's or designer's) perspective. We will focus discussions on applying the evaluation methodologies we learn to the AI and ML for games
techniques we read about in the published literature.
We will read papers from the literature in AI, ML, HCI, and games. Students will primarily be evaluated using a semester-long group project on a topic of their choosing (approved by the instructor). In addition, every student will be expected to present a paper from the AI or Game literature and lead a discussion about the design of an effective evaluation technique. There are no formal prerequisites; however, familiarity with the topics from a graduate-level artificial intelligence or machine learning course and/or a graduate level human computer interaction course are highly encouraged.
Design Graphics Lab
Dept. Computer Science
North Carolina State University
EBII 2280, 890 Oval Dr, Box 8206
Raleigh, NC 27695-8206