Artificial Intelligence

Course Information

Title:	Foundations of Artificial Intelligence
Institution:	Metropolitan State College of Denver
Semester:	Fall 2003 (August 18 - December 13)
ID, Section [CRN]:	CSI 4120, Section 1 [54726]
Meeting Times:	Tuesdays and Thursdays, 5:00 PM - 6:50 PM
Location:	Central Classroom 212
Course Website:	http://www.jodypaul.com/cs/ai
Discussion Board:	http://cs.mscd.edu/~discus
Instructor:	Dr. Jody Paul (schedule)
E-mail:	jody@computer.org
Office:	Science 133C
Office Hours:	Mondays/Wednesdays 6:50PM - 7:50PM Tuesdays/Thursdays, 1:00PM - 2:00PM Thursdays 9:30PM - 10:30PM Other days and times by appointment
Campus Mail:	Campus Box 38

Updates:

First Tri-mester Assignments [final versions due September 30]:
   Animal Game in Lisp/Scheme
   Tic-Tac-Toe using "Poor" Depth-First Strategy in Lisp/Scheme
   Tic-Tac-Toe using Minimax with Alpha/Beta Pruning in Lisp/Scheme
   First-Order Predicate Calculus / Resolution Theorem Proving exercise in Prolog
   Observations from Russel/Norvig text Chapter 10 (Knowledge Representation)
   Exercises from Russel/Norvig text -- 11.2, 11.5, 11.6, 11.9, 11.17, 14.12

Production System (simple Rule Interpreter & sample application) [due October 14]

Bayesian Network - Sample Problem [extra-credit]

Lecture Slides/Notes:
Neural Networks [Beaty] (PDF)
ART-1 [Paul] (PDF)

Script Applier Mechanism & English Language Interpreter:
MicroSAM (chez scheme), DrSAM (Dr. Scheme)
MicroELI (chez scheme), DrELI (Dr. Scheme)

Course Description:

This course is primarily a study of the theoretical foundations of artificial intelligence (AI), including the methods, techniques and tools used to investigate, design and build intelligent systems. Due to the vast breadth of the field of AI, a large number of topics will be surveyed, including: knowledge representation, machine learning, neural networks, cognitive science, intelligent agents, search strategies, resolution theorem proving, expert systems, reasoning under uncertainty, planning, natural language processing, and case-based reasoning.

Class sessions will consist primarily of discussions, critical analyses and design activities. A significant amount of preparatory reading and research throughout the course is thus required on the part of all participants.

Participants will develop demonstration prototypes using functional and logic programming paradigms and implemented in appropriate and commonly used AI programming languages (Prolog and LISP).

Resources:

Texts:

Artificial Intelligence: A Modern Approach (Second Edition)
     (Tattered Cover; Amazon)
by Stuart J. Russell & Peter Norvig
Prentice Hall (2003)
ISBN 0-13790-395-2
This is one big textbook. It covers a lot, but the authors have a bias that meant they left out significant parts of the AI field. Still, it's the best available right now. Be sure to get the Second Edition. Looks, feels and reads like a textbook. If you're ahead of the game, I'd read it a bit at a time; skip around and trade off with other books. (~1000 pages)

Understanding Artificial Intelligence
     (Tattered Cover; Amazon)
by the Editors of Scientific American; Sandy Fritz (Compiler)
Warner Books (2002)
ISBN 0-446-67875-9
A very accessible overview of AI. And it doesn't suffer from Russell & Norvig's bias. Probably good to read this book first. (~150 pages)

Dynamic Memory Revisited
     (Tattered Cover; Amazon)
by Roger C. Schank
Cambridge University Press (1999)
ISBN 0-5216-3398-2
A view from the "other" side of AI (missing from Russell & Norvig). This gives a good foundation for understanding how we may model memory and thought. You might never think about thinking the same way again. (~300 pages)

Can Animals and Machines Be Persons?
     (Tattered Cover; Amazon)
by Justin Leiber
Hackett Publishing Company (1985)
ISBN 0-8722-0002-7
This is an entertaining, thought provoking and very accessible presentation of deep issues in artificial intelligence and cognitive science. Excellent fodder for discussion.

Articles, Technical Reports, Journals & Conference Proceedings
Examples:
    IJCAI       AAAI       ACM SIGART       AI Magazine       AI-CBR
    Artificial Intelligence Bibliographies

Computing/Connectivity:
You must have World Wide Web access and an active e-mail account.
Note that you receive an e-mail account and Internet access by virtue of being a student at MSCD. (See: http://www.mscd.edu)
You are encouraged to make use of electronic mail to contact me often: jody@computer.org
You must have access to a computer that provides tools for document preparation and for authoring and editing graphics.
You will be expected to read and write programs using Lisp-derivative languages (such as Scheme) and Prolog. You are encouraged to take advantage of the extensive on-line information and tools.

Scheme, Lisp & Prolog Info:

  http://www.swiss.ai.mit.edu/projects/scheme/ (Scheme home)

http://www.scheme.org/ (Scheme links)
http://www.scheme.com/ (Chez Scheme)
http://www.plt-scheme.org/software/mzscheme/ (MzScheme)

Defining Description of Scheme (R5RS)
Lisp As an Alternative to Java, by Erann Gat, 1999 (PDF)
http://openmcl.clozure.com/ (Common Lisp for Linux & MacOS X)
http://sourceforge.net/projects/jlogic/ (JLog - Prolog in Java)
http://sourceforge.net/projects/gprolog/ (GNU Prolog)

Grading Policy:

You are expected to make several in-class presentations and to participate in class discussions and in-class exercises. There will be homework assignments that you are required to complete and turn in. Your final course grade is determined by combining scores on the exercises, presentations, and assignments. You are guaranteed a grade no lower than that given by the following conversion of score (percentage of total possible) to letter grade:
89% < A; 79% < B < 90%; 69% < C < 80%; 59% < D < 70%; F < 60%

N.B.: Participation in class discussions and exercises is mandatory.

Late assignments will not earn course credit. You may submit an assignment after its due date for comments and advice, and you are encouraged to do so. However, the score for that assignment will be recorded officially as 0. Likewise, missing an in-class exercise will result in a score of 0 for that exercise. Specifically, late homework and make-up exercises will not be accommodated without prior arrangement and written agreement. Unforeseeable crises and emergency situations will be dealt with on a case-by-case basis in accordance with MSCD, College, and Departmental policies.

Note that a substantial amount of information will be disseminated during class sessions or on course websites that you will be responsible for knowing whether or not you attended the sessions or accessed the website. Note in particular that the textbooks do not provide all of the information necessary to successfully complete the assignments and exercises.

Collaboration

I encourage collaboration and regard it as essential aspect of Computer Science. Collaboration and discussion with fellow students concerning course information, materials, proofreading, concept exploration, and studying for exams is encouraged. You are not expected to learn the course content or work on assignments in a vacuum on your own. However, you must write up your own solution, individually, to every assignment you turn in even if the solution results from a collaborative effort. In your write-up, you must credit the people with whom you worked. If you consult any reference material, please note in your assignment which sources you used for each part. Note that collaboration is not acceptable during any exam. Turning in work that is the result of copying, failure to credit your collaborators, lack of citations for references, and attempts at collaboration or copying during exams will be treated as academic dishonesty. All incidents of suspected dishonesty will be reported to the department and the Dean of the college. Consequences may include a grade of 0 on the assignment or exam, a grade of "F" for the course, academic probation, or dismissal from the institution. This is a very serious matter and should not be taken lightly. If you have any uncertainty or concerns, please discuss them with your instructor or advisor.

Official Announcements:

Important Dates and Deadlines:

See MSCD Calendar

Class Attendance on Religious Holidays:

The college policy on Class Attendance on Religious Holidays is posted on the information board outside the Mathematical and Computer Sciences department office (SI141). In addition, copies of this policy are available from the department upon request. It is the students' responsibility to understand and abide by the policy.

American with Disabilities Accommodations:

Students desiring a reasonable accommodation under the ADA must contact the instructor immediately to discuss their needs. Failure to notify the instructor, in a timely manner, of the need for a reasonable accommodation may hinder the college's ability to assist students in successfully completing the course.