Course Description
Choosing optimally among different lines of actions is a key aspect of autonomy in artificial agents. This course will focus on how to make optimal and approximately optimal decisions in single and multiagent settings. It will be self-contained, introducing background literature such as aspects of probability and game theories.
Athena Title
DECISION MAKING
Prerequisite
CSCI 4470/6470 or permission of department.
Semester Course Offered
Not offered on a regular basis.
Grading System
A - F (Traditional)
Course Objectives
1. To study algorithms for automated decision making by agents situated in uncertain single-agent and multiagent environments 2. To understand key subareas of artificial intelligence, economics and cognitive psychology 3. To become proficient in the use of computing tools related to decision making, and designing and giving presentations effectively
Topical Outline
I. Introduction a. Requirements for decision models and solutions b. Probability theory background c. Bayesian networks and Influence diagrams II. Decision making in single agent setting a. Markov decision processes (MDP) b. Partially observable Markov decision processes (POMDP) c. Dynamic influence diagrams d. Psychology of decision making III. Decision making in multiagent setting a. Game theory background b. Repeated strategic and Bayesian games c. Decentralized MDPs d. Partially observable Dec-MDPs (Dec-POMDPs) and approximations e. Interactive POMDPs (I-POMDPs) and approximations f. Interactive Influence diagrams
Syllabus