cse250ahw3

cse250ahw3 - CSE 250a. Assignment 3 Out: Tue Oct 12 Due:...

Info iconThis preview shows pages 1–3. Sign up to view the full content.

View Full Document Right Arrow Icon
CSE 250a. Assignment 3 Out: Tue Oct 12 Due: Tue Oct 19 3.1 Node ordering Recall the alarm belief network (BN) described in class, with node ordering { B,E,A,J,M } and directed acyclic graph (DAG) shown below: B J A M E Draw the (minimal) DAGs that would be required to represent this same joint distribution for the following alternative node orderings: (a) { A,J,M,B,E } (b) { A,B,J,E,M } (c) { B,M,J,A,E } (d) { B,A,J,E,M } (e) { J,B,M,A,E } (f) { J,A,B,E,M } It is not necessary to compute the CPTs in each of these alternative BNs. However, for each DAG, you should briefly justify the edges that you include or omit by appealing to properties of conditional indepen- dence. 1
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
3.2 Stochastic simulation Consider the belief network shown below, with n binary random variables B i ∈{ 0 , 1 } and an integer random variable Z . Let f ( B ) = n i =1 2 i - 1 B i denote the nonnegative integer whose binary representation is given by B n B n - 1 ...B 2 B 1 . Suppose that each bit has prior probability P ( B i =1) = 1 2 , and that P ( Z | B 1 ,B 2 ,...,B n ) = ± 1 - α 1 + α ² α | Z - f ( B ) | where 0 < α < 1 is a parameter measuring the amount of noise in the conversion from binary to decimal. (Larger values of
Background image of page 2
Image of page 3
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 11/20/2010 for the course CSE 250a taught by Professor Lawrence during the Spring '10 term at UCSD.

Page1 / 3

cse250ahw3 - CSE 250a. Assignment 3 Out: Tue Oct 12 Due:...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online