SCXT
350
Take-home
final exam
Fall
2000
General
notes:
1, Responses should be typed (word processor) except where a
diagram is necessary.
2. Email submissions can not be accepted.
3. You are welcome to consult published works (be sure to cite
them), but you are explicitly prohibited from consulting with fellow students
(even to clarify the meaning of a problem).
This must be completely and entirely your own work. Do please feel free to ask me questions, of
course.
4. Problems should (once again) be typewritten (word
processor), and should be submitted in order.
It is always a good idea to
retain a copy for yourself.
.5. This exam will be turned in at 10:00 AM Friday, Dec. 15.
I will be in our classroom
from 9:30 AM to 10:15 AM to collect exams.
There
have been several threads to our discussions this term. In this exam, we visit them one more time. Each of the following will have roughly
equal weight, and I would expect roughly two pages (more than one, no more than
three) for each.
1. Physical Symbol System Hypothesis: Intelligent action is computable.
Consider the Descartes, Turing, Newell and Simon, and
Searle papers in the coursepack to be a time-constrained discussion starting
with Descartes and continuing through Searle.
In the form of an introduction to the coursepack, outline (in essay
form, of course) the main issues raised in this conversation. Be sure to address in your introduction (in
addition to other issues you may wish to discuss) the notion that intelligent
action is computable. In a final
paragraph (that might lead to the inclusion of another paper in the
coursepack), address briefly the distinction between symbolic and connectionist
AI. Important Note:
Students in CSci 431 who have started work on problem 6 (b) should pick
another option for problem 6.
2. Computation.
Beginning with the simple model of a computer we
discussed at the very start of the class, describe the various models of
computation we have discussed this term.
Your discussion should include this first model, programming in LISP, finite
state automata, Turing machines, and the physical symbol system. In your discussion of the physical symbol
system, be sure to incorporate a discussion of search and knowledge
representation as we have discussed it in class.
3. Marr
Marr gives us a way to organize our thoughts on complex
phenomena. Give a description of Marr's
three levels as you would describe them for an intelligent friend. In your discussion, consider specifically either
problem-solving or natural language understanding (please be sure to re-read
Marr's paper in either case), and incorporate examples from our discussion in
the course on knowledge representation, search, grammars, and expert systems.
4. Other issues. Do
one of the following (one of which would involve a bit of research).
a. Consider the paper on expertise and also our work with
expert systems. Give descriptions of
both, and then discuss how they are related and how they fail to relate.
b. Consider either the lecture on memory (Professor Mark
Reinitz) or the lecture on the development of the facility in children for
understanding and producing natural language (Professor Cathy Hale). Review (in written form) the basic parts of
the lecture (and associated article), and explore how these have been modeled in
artificial intelligence. Be sure to
cite any references you use.
Please
Note: Although this is given as a
draft, I do not intend to make any major changes to it. Any changes (unless I find that I've really
done something wrong) will be on the order of clarifications. No changes are anticipated after today.