Computational Models of the Mind

Fall 1996

Instructor

Craig Miller
256 South College, x-1400
Email: millercr
Office hours to be announced and placed on Web page

Meetings

TuTh 9:30-11
Rm 1 (basement) South College

Text

The Computer and the Mind, by Philip N. Johnson-Laird.

Overview

This course introduces the area of cognitive science, an interdisciplinary field that draws upon aspects of cognitive psychology, philosophy, linguistics, cognitive anthropology, as well as computer science. Its unifying goal is to construct and evaluate process-oriented theories of how people think and reason. Cognitive science research typically works under the assumption that human thinking processes can be understood in computational terms.

The course takes a strong computer science perspective. You will be working and experimenting with actual computer programs that purport to be models or simulations of some aspect of human cognition. In some cases, your goal will be to modify these programs and then re-evaluate them as a new scientific theory. We will discuss to what extent these models constitute an explanation for how people think as well as debate whether certain models serve as viable scientific theories.

Goals

Certainly, one goal is to learn about cognitive science, and even about science in general. But the course will also target research and professional skills that have a broad utility in almost any computer-oriented field. These skills include:

Much of the course will take on a seminar format where you will be responsible for reading the assigned material before class and then come prepared to discuss it. Occasionally, lectures will address some of the more difficult concepts and supplement the material in the text.

Projects

Projects will also play an important role in the course. You may find that they are generally more open-ended than what you have experienced in previous computer science courses. While an assignment will typically start you with a specific task, you will be required to take your own initiative in expanding on the assigned work in useful and interesting ways. Despite the open-endedness, you will often receive suggestions in guiding you on this process. Ultimately, you are required to write up what you have done, describe and analyze your results, and present any conclusions drawn from them. Because of the open-ended nature of the projects, you will receive intermediate feedback on most of your projects and then have the opportunity to revise them.

Tentative Project Schedule

Project Due Date Rev. Date
Verb Generation Sep 12 Sep 24
Neural Net Learning Oct 8 Oct 22
Symbolic Concept Learning Oct 31 None
Subtraction process Nov 14 None
Advanced Project Dec 5 Dec 17

Turning in a revised project is not mandatory but strongly encouraged.

The midterm will be held Tuesday Oct. 15.

The final is scheduled for Friday Dec 13, 2pm.

Grade Determination

40% Projects (4)
15% Advanced project
15% Mid term
15% Final
15% Preparation, quizzes, participation

Tentative Schedule

Day Topic Reading
Aug 29 Course intro Handout
Sep 3 Scientific theories Ch. 1
5 Cognitive Processes Ch. 2-3
10 Architecture Overview TBA
12 Perception Ch. 4-5
17 Perceptual representation Ch. 6
19 Learning intro Ch. 7
24 PDP Learning Ch. 10
26 More PDP Learning Ch. 10
Oct 1 Memory Ch. 8
3 Induction and concepts Ch. 13
8 Plans and productions Ch. 9
10 Production systems Handout
15 Midterm --
22 Rule-based classification --
24 Deduction Ch. 12
29 Creation Ch. 14
31 Project #3 discussion Category Learning Data
Nov 5 Project #4 presentation Brown & VanLehn Article
7 Language Ch. 15 & 17
12 Meaning Ch. 18
14 Article discussions Conference Articles
19 Writing project abstracts --
21 Work on projects --
Dec 3 Work on projects --
5 Project presentation --
10 Consciousness Ch. 19-20