CS 540 - Introduction to Artificial Intelligence
Spring, 1997
Office: 6385 CS & Stats Building
Telephone: 263-2874
Email: kunen@cs.wisc.edu
Office Hours: Fridays, 9 - 11, or by appointment, or by email.
Login: kunen
TAs
NAME OFFICE HOURS LOGIN
Bryan So 5364 CS&S Tuesday, 1-2 so
Xin Wang ERB 735 Thursday, 2:30-4:30 xwang
Igor Ivanisevic 1304 CS&S M and W, 10-11 iigor
Grades
There will be four programming assignments, counting 10% each, and
three exams, counting 20% each; the third exam will be at the
scheduled time and place for the final. Programs may be turned in
any time (until midnight) on the day due. Late assignments will
loose 5% for the first day late, and 10% for each day late after
that. No assignments will be accepted after May 11.
Schedule
- Program 1 due: Thursday, February 20. click here to see it on line; it is also in the
course directory as proghw1.ps
- Program 2 due: Thursday, March 20. click
here to see it on line; it is also in the course directory as
proghw2.ps
- Program 3 due: Thursday, April 17. click
here to see it on line; it is also in the course directory as
proghw3.ps
- Program 4 due: Thursday, May 8. click
here to see it on line; it is also in the course directory as
proghw4.ps
- Exam 1: Thursday, February 27.
- Exam 2: Thursday, April 3.
- Exam 3: Sunday, May 11, Mothers' Day, 7:25 - 9:25 PM, in 180
Science Hall; that's the old red brick building, just southwest of
Memorial Union.
Topics Covered
The following order of topics isn't entirely logical, but is
designed so that the topic for each programming assignment is
discussed before the program is due.
- Introduction to Lisp (Program 1).
- Searching and Game Playing (Program 2). Text 3,4,5
- Introduction to Prolog. Text 10.3
- Natural Language Understanding (Program 3). Text 22,23
- Learning and Neural Networks (Program 4). Text 19
- Logical Deduction. Text 9, 10.2, 10.4
- Planning. Text 11, 12
- Reasoning with Uncertain Knowledge. Text 14,15
Chess
Click here for information
about the Deep Blue -- Kasparov chess rematch.
Automated Reasoning Information
The program OTTER is the best general-purpose resolution-based
theorem prover available. Click here for
more information, source code, etc.
The course directory contains the executable code (for the
instructional machines), and some sample otter input files (of form
*.in).
The usual call to otter is:
otter < foo.in >! foo.out
Neural Nets Information
Click here to see a simple version of
backpropagation, sufficient for the fourth programming assignment.
The following COURSES will be given next semester (Fall, 1997),
and contain material relevant to neural nets:
Lisp Information
Since Lisp is used in a lot (not all!) of AI programming, the
course will begin with a discussion of COMMON Lisp. It would
probably be useful to have some Lisp reference available to
supplement the lectures and the on-line help available within lisp
itself. There are many paperbacks available, most of which are
probably ok.
- I like Common LISPcraft by Wilensky.
- Another possibility: The ANSI Common Lisp Book by
Graham. Code used in the
book is on line.
- The Ultimate Lisp Reference: Steele's Common Lisp:
The Language (2nd Edition) , 1029 pages. Also
available on line.
- Also:
Harlequin Books maintains a hyper-text lisp reference.
- Click here
for more information on using Common Lisp on the Suns
- Click here
for more information on the Gnu Common Lisp implementation
Prolog Information
Information about Prolog is available from the following sources:
- The text, Section 10.3.
- There are many paperback introductory manuals around. I like
the ones by Clocksin - Mellish, and by Bratko.
- The "help" facility in Quintus Prolog gives you an index into
the entire Quintus Prolog manual.
- Two sample programs are in the course directory: db.pl is an
example of a deductive database program, and grammar.pl shows how
to use Prolog for natural language processing.
- The deductive database language Coral has a syntax similar to
Prolog's, and is preferable to Prolog for many database tasks.
Additional Information
- For information about the AFS file system, click
here.
- To see pseudo-code for alpha-beta search, click here.
- To see pseudo-code for best-first search, click here.
- An alpha-beta problem from a previous exam is on the course
directory (alpha_beta.ps). click here
to see it on line.
- Textbook: Artificial
Intelligence: A Modern Approach by Russell and Norvig
- Class time: 8:30 - 10:45 TR (Lec. 1) and 11:00 - 12:15 TR (Lec.
2) in 1325 CS.
- Recitation sessions: 2:30 - 3:45 T in 2305 Engr (Lec. 1) and
4:00 - 5:15 T in 121 Psych (Lec. 2). In these, no essentiallly new
material will be presented. I'll answer questions, give hints for
programming assignments, and review for exams. They will usually
last only 50 minutes. Since I am teaching both sections, you may
attend a recitation section which is different from your lecture
section.
- The course directory is /p/course/cs540-kunen/public
- A best-first-search problem from a previous exam is on the
course directory (astar.ps). click here to
see it on line.
Exams from Fall, 1995 (postscript)
Last Changed: April 25, 1997 by
kunen@cs.wisc.edu