Practical Probabilistic Modeling with Graphical Models

EE 639 Advanced Topics in Signal Processing and Communication

Fall 2009  


.

[Home] [Lectures] [Homework]


Dates

Lecture

8/26

Reading: “Graphical Models” by M. I. Jordan (focus on applications)

Introduction to the course (updated: 8/28)

8/28 – 9/14

Reading: Chapter 2

Preliminaries (update: 8/31)

Parameterization of Joint Probability

Directed Graph Representation

DAG Conditional Independence

Bayes Ball Algorithm

Undirected Graph Representation

9/16

Reading: Chapter 3

The Elimination Algorithm

9/21-9/23

Reading: Chapter 4

Probability Propagation and Factor Graphs (1) (2)

9/28-10/5

Application: Error Control Coding and Loopy Belief Propagation

 (1) (2)

10/7 – 10/12

Reading: Chapter 5

Statistical Concepts (1) (2)

10/14 -10/21

Reading: Chapter 6

Linear Regression and the LMS Algorithm (1) (2) (3, video)

10/26-10/28

Reading: Chapter 7, 8

Linear Classification (1) (2) (3)

The exponential family and generalized linear models

11/2 – 11/4

Reading: Chapter 9, 10

Completely Observed Graphical Models (1)

Mixtures and conditional mixtures (1)

11/9

Reading: Chapter 11

Expectation Maximization (1)

11/16

Reading: Chapter 12

Hidden Markov Models (1) (2)

Application: Speech Recognition and Synthesis

11/16 – 11/18

Reading: Chapter 13, 15, 18

The Multivariate Gaussian

Kalman Filtering

The HMM and State Space Model Revisited

Application: Object Tracking

11/23 – 11/30

Reading: Markov Properties

Reading: Chapter 17

Junction Tree

Skipped?

Reading: Chapter 19

Features, maximum entropy, and duality

Skipped?

Reading: Chapter 20

Iterative scaling algorithms

12/2

Reading: Sampling Methods

Application: Object Tracking with Particle Filters

12/4

Final project poster presentation

12/7 – 12/11

Reading: Graphical models, exponential families and variational inference

Application: Image Segmentation

 


Sen-ching Samson Cheung

 Last update: 8/24/2009