CMPS 287 Final, Fall 2001
American University of Beirut · Artificial Intelligence · 1 page
Stuck on a question? A tutor can walk you through it.
More CMPS 287 files
Show the text of this final exam
=
=
ey
5
AMERICAN UNIVERSTT4
LIBRARY
OF BEIRUT
American University of Beirut
Finaly sath, 287- Fall 01-02
Pf. Jana Sukkarieh
February 8th, 02
It is an Open notes Open books exam as you know. It is an open-ended final.
Use your own judgement (a good one) on how detailed your answer should be.
Your intelligent agents- the ones you developed for your projects-
played games. organised/controlled ‘appliances’ in our homes, rea-
soned with set theory, recognised language, processed English, or
sniffed the "smell" of intruders, etc .. Your research papers may have
covered even more intelligence!
You all claimed that you have used techniques present in Norvig and
Russell ... that you've used some AI technique or another ...
You all have reasoned under certain or Aristotelean environment in
the sense that vour knowledge representation was not. probability
theory!
Twant you, in this final, to investigate into changing your intelligende
agent into one that is capable of reasoning under uncertainty.
1. Define your rules and knowledge base in English
2. Transform them into a suitable representation (probability the-
ory language!)
3. Use Belief networks (or maybe some other proof theory) to
reason
4. etc
I leave it to yon to explore into the capabilities and limitations of
your own “baby” i.e. your project-designed agent.
Note that people who submitted the same project should work on different
domains or work on different angles of the reasoning process(in any case I have
expected double work for your project and research paper).
Bonus: You cau investigate into making it a learning agent. (if you haven’t done
so in vour project).
Best of luck
OF BEIRUT
TEE