Scribd
Upload
22 views

Assignment+4 1

Uploaded by

Praveen Jayramraj
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
22 views

Assignment+4 1

Uploaded by

Praveen Jayramraj
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Assignment 4

Written Assignment: Uncertainty


and Bayesian Network
TASK 1 (20 points)
Consider the given joint probability distribution for a domain of two variables (Color, Vehicle):

Color = Red Color = Green Color = Blue


Vehicle = Car 0.1299 0.0195 0.0322
Vehicle = Van 0.1681 0.0252 0.0417
Vehicle = Truck 0.1070 0.0160 0.0265
Vehicle = SUV 0.3103 0.0465 0.0769

a) (10 points) Calculate P (Color is not Green | Vehicle is truck)


b) (10 points) Prove that Vehicle and Color are totally independent of each other
TASK 2 (30 points)
a. (10 points) Function P1 is a function defined on a set of samples S = {A, B, C, D}. P1 is defined as
shown below. Is there a value for y that makes P1 a valid probability function? If yes, what is that
value? Justify your answer.

P1(A) = 10*y

P1(B) = 20*y

P1(C) = 30*y

P1(D) = 20*y

b. (10 points) Function P2 is a function defined on the set of real numbers. P2 is defined as shown
below. Is there a value for y that makes P2 a valid probability density function? If yes, what is that
value? Justify your answer.

P2(x) = 0 if x < 100

P2(x) = 6*y if 100 <= x <= 110

P2(x) = 4*y if 110 <= x <= 130

P2(x) = 0 if x > 130.

c. (10 points) Function P3 is a function defined on the set of real numbers. P3 is defined as shown
below. Is there a value for y that makes P3 a valid probability density function? If yes, what is that
value? Justify your answer.

P3(x) = 0 if x < 0

P3(x) = 7*y if 0 <= x <= 10

P3(x) = 3*y if x > 10


TASK 3 (30 points)

A P(B|A) B P(C|B)
P(A) T 0.8 T 0.2
0.6 F 0.3 F 0.9

A B C

a. (15 points) Given the above Bayesian network, compute P( (B=true) AND (C=false) ). Write the
complete expression and the final numerical answer

b. (15 points) In the above Bayesian network, compute P(B).Write the complete expression and the
final numerical answer

TASK 4 (20 points)


a (10 points). You have a Bayesian network of N nodes. Each node corresponds to a Boolean
random variable. Each node has a maximum of 3 parents. How many numbers would you need to
specify at most, in order to fully specify the probability distribution modeled by this Bayesian
network? In other words, what is the maximum number of values you need to store (for the entire
network) in order to full specify the probability table for each node? Justify your answer.

b (10 points). Suppose that you want to model the same probability distribution as in question a
but using a joint distribution table. How many values do you need to specify in that case?

You might also like