COMPSCI 683: Artificial Intelligence

This is a brief overview of the course syllabus and assessment policy, for students who are considering taking this class. More details are available on Moodle.

• Overview

  This is an introductory class to artificial intelligence. It's a really fun class (I think!) since it describes the fundamentals of some important AI systems. In particular, we'll explore the basics of how AI systems explore and reason about the environments they operate in. So:

  • Searching for solutions when there's no/partial information, and when you are playing against an adversary who wants to beat you.
  • Constraint Satisfaction Problems: how can you find a feasible solution under constraints.
  • Planning Under Uncertainty: how can an agent model its goals and take actions in order to reach them, in an uncertain (or even adversarial) environment.
  • Logical Agents: how can agents model their world using logic, and how can we infer new knowledge based on what we already know.
  • Bayesian Inference: just like logic, only better! With uncertainty and relationship between variables, we'll have fun with probabilities!

  This class is going to be focused on theoretical analysis, expect little programming and more theorem proving!

• Preparation for Class

  There is a strong correlation between poor performance in my classes and inadequate preparation. Students that do not invest sufficient time preparing and submitting assignments tend to do poorly. This is not a class that you can cram for. My classes tend to focus on theoretical analysis, and you need to be comfortable writing mathematical proofs. Practice is essential, especially if you have no prior experience in writing rigorous proofs. I strongly encourage you to attend lectures. While recordings will be available via Moodle, they are an insufficient substitute for regular attendance. If you believe you won't be able to regularly attend lectures, please let me know ASAP. More generally, if you see that you are falling behind it is always better to discuss things with the course staff before the course ends. Resolving issues after the fact is almost always more difficult than anticipating and dealing with them before the final exam.

• Assessment

  Course assignments account for 50% of your grade. We will also have two midterms, each is 25% of your grade. There is no final exam. I will offer bonus points via additional (extra fun!) homework problems, and participation in class.