Probabilistic Graphical Models 10-708
Homework 5 (Extra Credit): Due April 30, 2014 at 4 pm
Directions. Note that this homework is only two problems, both of which are extra credit. Hence,
if you do not do this homework, your grade will not be penalized.

School of Computer Science
Probabilistic Graphical Models
Directed GMs: Bayesian Networks
Eric Xing
Lecture 2, January 15, 2014
Receptor A
X1
Kinase C
X3
Receptor B
Kinase D
TF F
Gene G
X7
X4
X2
Kinase E
X5
X6
Gene H
X8
Reading: see class homepage
Eric X

School of Computer Science
Probabilistic Graphical Models
Maximum likelihood learning of
undirected GM
Eric Xing
Lecture 8, February 10, 2014
Reading: MJ Chap 9, and 11
Eric Xing @ CMU, 2005-2014
1
Undirected Graphical Models
Why?
Sometimes an UNDIRECTED

Probabilistic Graphical Models 10-708
Homework 4: Due April 14, 2014 at 4 pm
Directions. This homework assignment covers the material presented in Lectures 13-18. You
must complete all four problems to obtain full credit. To submit your assignment, please

Probabilistic Graphical Models 10-708
Homework 1: Due January 29, 2014 at 4 pm
Directions. This homework assignment covers the material presented in Lectures 1-3. You must
complete all four problems to obtain full credit. To submit your assignment, please

Probabilistic Graphical Models 10-708
Homework 2: Due February 24, 2014 at 4 pm
Directions. This homework assignment covers the material presented in Lectures 4-8. You must
complete all four problems to obtain full credit. To submit your assignment, pleas

Probabilistic Graphical Models 10-708
Homework 3: Due March 21, 2014 at 4 pm
Directions. This homework assignment covers the material presented in Lectures 9-12. You must
complete all four problems to obtain full credit. To submit your assignment, please

School of Computer Science
Probabilistic Graphical Models
The Belief Propagation
(Sum-Product) Algorithm
X1
Eric Xing
m21(x1)
Lecture 5, January 29, 2014
m32(x2)
X2
X3
m42(x2)
Reading: KF-chap 10
X4
Eric Xing @ CMU, 2005-2014
1
Pros and Cons of
Procedure

School of Computer Science
Probabilistic Graphical Models
Factor Analysis and State Space
Models
Eric Xing
Lecture 11, February 19o, 2014
Reading: See class website
Eric Xing @ CMU, 2005-2014
1
A road map to more complex
dynamic models
discrete
Y
discret

School of Computer Science
Probabilistic Graphical Models
Gaussian graphical models and
Ising models: modeling networks
Eric Xing
Lecture 10, February 17, 2014
Reading: See class website
Eric Xing @ CMU, 2005-2014
1
Where do networks come from?
The Jesus

School of Computer Science
Probabilistic Graphical Models
Representation of undirected GM
Eric Xing
Lecture 3, February 22, 2014
Reading: KF-chap4
Eric Xing @ CMU, 2005-2014
Two types of GMs
Directed edges give causality relationships (Bayesian
Network o

School of Computer Science
Probabilistic Graphical Models
Exact Inference:
Variable Elimination
B
A
A
B
md
Eric Xing
A
mb
mc
C
A
C
me
E
G
A
mf
D
A
mh
E
E
Lecture 4, January 27, 2014
F
D
C
mg
E
Reading: KF-chap 9
F
H
Eric Xing @ CMU, 2005-2014
1
Recap:
De

School of Computer Science
Probabilistic Graphical Models
Introduction to GM
and
Directed GMs: Bayesian Networks
Eric Xing
Receptor A
X1
Kinase C
Receptor B
X3
X2
Lecture 1, January 13, 2014
Kinase D
TF F
Gene G
X7
Kinase E
X4
X5
X6
Reading: see class hom