Assignment 3 (Due 9/8, Wed, 6pm):
Assignment:
- Read Box 3C (on page 64) and sections 3.3, 3.4.1, 3.4.2, and 3.5
from Chapter 3.
- Turn-in at least
two questions from the above reading.
- Design and turn-in a Bayesian network structure of at least five variables for
a domain of your choice. More specifically, write down the random
variables, and the links between the random variables. Use your intuition (not tools). There is no need
to specify the parameters. Explain in at least three sentences
what the domain is and what the interactions are. For example:
The Student Example:
Random variables: Intelligence, Difficulty, SAT, Grade, Letter
Edges: Intelligence -> SAT, Intelligence -> Grade, Difficulty
-> Grade, Grade -> Letter
Explanation: In this domain, I modeled a typical undergraduate
admission domain. In this domain, the intelligence of the student and
difficulty of a class affect his/her grade, which in turn affects the
quality of the recommendation letter. His intelligence also affects his
SAT score.
Note: You cannot use
this example :)
Instructions:
- Turn-in your assignment (the questions and the bayesian network)
by email, w/ the subject: CS595 - Assignment # 3
- Everything should be written in plain text. No .doc, .pdf, etc
files, please.