Schedule of Topics, Tasks, and Exams

This is a plan. As you will learn, planning has been a central topic in AI for decades. It's not easy and it's fraught with error. Personally, I'm with Eisenhower: "Plans are worthless, but planning is everything."

So, here's the plan. There will be 3 projects in this class: the Solver, the Thinker, and the Learner.

The Solver will use constraint satisfaction problem solving (CSP) techniques. CSP is a very general framework for posing and solving problems in terms of assigning values to variables.
The Thinker project will use logical knowledge representation and inference techniques to solve problems and reason about the world.
The Learner project will use some basic machine learning techniques to develop a problem solver that learns.

Each project will take approximately 1/3 of the course. The last class is June 3. Allocating 3 weeks per project, we get the following estimated timeline:

Project	Due Date
Solver	April 15
Thinker	May 6
Learner	May 27

These projects meant to be interesting, somewhat realistic, focused on core classic AI problems, and challenging. They are not doable two days before they are due.

Based on requests in the CTECs, the quizzes will be the week after the projects are due.

Exam	Date
Quiz 1	April 20
Quiz 2	May 11
Final	June 7, 3-5pm

The concepts and algorithms for each project are described in the textbook. The quizzes and final exam will include some of this material but focus more on things from the book not included in the project. Quizzes and exams are closed book.

Theme	Readings
Search-based Problem Solving	Chapters 3, 4, 5 Skippable: Section 3.6, Memory-bounded Search, Section 4.3-4.4 (but Fig 4.10 is important)
Logic-based Inference	Chapters 8, 9, 10 Skippable: Circuits example in Chap 8, Sec 9.5, Sec 10.7-10.8
Planning, Learning	Chapters 11, 12, 18, 19

Note that the third section has two major topics, only one of which will be included in a project.

Comments?

Send mail to Chris Riesbeck.