======================================================================

CS 4611 and CS 7611
AI Problem Solving
Spring 2005

General Information

----------------------------------------------------------------------

Meeting Place: CoC 101
Meeting Times: 11:00 to 12:00 am, Mondays, Wednesdays and Fridays

Instructor: Ashok Goel 
Email: goel@cc.gatech.edu
Phone: 404-894-4994
Office: CRB 394 - however note that office hours will be held in the 
        CCB first-floor lounge
Office Hours: 10:00 - 11:00 am, Mondays, Wednesdays and Fridays - note
        that the best way to contact me at other times is by email.

TA: TBS


----------------------------------------------------------------------

Description:
  CS 4611/7611 AI Problem Solving is a 3-credit joint undergraduate/graduate 
  course. The pre-requisites are CS 4600 (Intelligent Systems), or CS 6601
  (Artificial Intelligence), or equivalent. Undergraduate and graduate
  students will be graded separately (i.e., their grades will be normalized
  only with respect to their peers). Also, only graduate students will be 
  required to do a term project as part of the course work. 

Format: 
  The course will consist of class lectures and discussions, a series of
  reading assignments, three design and programming projects, and two
  examinations.  In addition, graduate students will be required to do
  a term project. 

----------------------------------------------------------------------

Objectives:
  A better title for this course might be knowledge-based AI, which
  includes reasoning, learning and problem solving. Its called AI 
  Problem Solving because the examples and exercises will be from 
  problem solving (instead of language, perception of action).

  The objectives of the course are to introduce the basic concepts, models,
  methods and issues in knowledge-based AI. By the end of the course, a 
  student should know enough about knowledge-based AI to (i) take the relevant
  portions of the Ph.D. qualifying examination in Intelligent Systems, and 
  (ii) conduct research with any of the Intelligent Systems faculty working 
  in knowledge-based AI. 

Scope:
  The course will cover knowledge-based reasoning, learning and 
  problem solving. Specific methods will include conceptual reasoning, 
  rule-based reasoning, case-based and analogical reasoning, qualitative 
  modeling and model-based reasoning, visual reasoning, and meta-reasoning.

  The course will also cover knowledge representation and organization.
  Specific topics will include production rules, discrimination networks, 
  semantic networks, frames and scripts, and models of physical systems, 
  information systems, and software agents.

  Specific problems will include classification, diagnosis, configuration,
  design, and discovery. 

----------------------------------------------------------------------

Book: 
  There is no prescribed textbook.  Much of the material in the
  first month of the semester will be from Artificial Intelligence by
  Patrick Winston, MIT Press, 1993, 3rd Edition. Similarly, much of the
  material in the second month will be from Knowledge Systems by Mark
  Stefik, Morgan Kauffman, 1995.  Copies of readings will be handed out
  in the class as and when needed.

Webpage and Swiki: 
  The TA will maintain a webpage for the class, which will provide 
  information about the course including a day-by-day class schedule and 
  reading assignments. He will also maintain a class swiki, which will 
  provide a forum for posting class notes and discuss related issues. 

-------------------------------------------------------------------------

Classnotes:   
  Students will take turns as "designated note-takers," and post them to
  class swiki. The notes for a given class must be posted by the next class.

Design and Programming Projects:
  Each students will do three design and programming projects, where 
  the second and third projects will build on preceding ones. Also, the
  projects will become increasingly open-ended, with the first one being
  mostly circumscribed and the third one largely open-ended. 

Term Project:
  Each graduate student will do a term project which must have design and 
  programming components. 

Examinations: 
  Both examinations will be open-notes. The mid-term examination will come 
  roughly in the middle of the semester. The final exmination will be 
  comprehensive (i.e., it will cover the entire course).

----------------------------------------------------------------------

GRADING: 
  Undergraduate and graduate students will be graded separately. 
  For UGs, each of the three short projects will count for 12% of the grade, 
  the mid-term for 20% and the final for 40%, with 4% reserved for class 
  attendance and participation. 
  For Grads, each of the three short projects will count for 9% of the grade, 
  the mid-term for 15% and the final for 30%, and the term project for 25%; 
  3% is reserved for class attendance and participation.

----------------------------------------------------------------------

Ashok Goel
January 4, 2005

======================================================================