CSC2542: Topics in Knowledge Representation and Reasoning:
AI Planning and Search
Winter 2021  


Instructor: Sheila McIlraith , Pratt 398D, csc2542profs -at- cs or sheila -at- cs
TA: Maayan Shvo, csc2542tas -at- cs

Timing: Wednesday, 3:00 - 5:00 EST
First class: Wednesday, January 13, 2021
Access: Quercus

Office Hours: Timing to be determined.

Questions about course access: Please email csc2542tas-at-cs

Course Description: Automated planning is a subfield of AI that concerns the generation of plans, policies, or strategies for execution by agents. It is central to sequential decision making and to the construction of intelligent agents and autonomous robots. In this course, we will investigate theory and algorithms for automated plan generation and execution with and without the existence of reliable models, with a focus on search and learning. The theory and algorithms we will be exploring in this course are applicable to a diversity of problems beyond the development of intelligent agents or cognitive robots, including activity and plan recognition, program synthesis, automated monitoring and diagnosis, and software and hardware verification.

Topics to be covered include: search techniques, automated planning, MDPs, reinforcement learning.

Required preparation: CSC384 (Introduction to Artificial Intelligence) or comparable.

Announcements


January 11 Class Logistics
  • This course will be run as a synchronous course.
  • Access is via quercus.
  • First class is Wednesday, January 13, 2021 at 3:00 PM (we start at 3:10pm)
  • Lectures will be delivered via zoom. Please make sure to login with UofT credentials.
  • Please write to the TA csc2542ta -at -cs if you have issues connecting.