Download presentation
Presentation is loading. Please wait.
1
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 1 Jeff A. Bilmes University of Washington Department of Electrical Engineering EE512 Spring, 2006 Graphical Models Jeff A. Bilmes Lecture 2 Slides March 30 th, 2006
2
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 2 d-separation, 3 canonical BNs, Bayes ball Undirected Models Start computing probabilities Outline of Today’s Lecture
3
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 3 Books and Sources for Today Jordan: Chapters 1 and 2 Derin 1989 (Markov Random Fields) Lauritzen, 1986. Any graph theory text.
4
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 4 L1: Tues, 3/28: Overview, GMs, Intro BNs. L2: Thur, 3/30: semantics of BNs + UGMs L3: Tues, 4/4 L4: Thur, 4/6 L5: Tue, 4/11 L6: Thur, 4/13 L7: Tues, 4/18 L8: Thur, 4/20 L9: Tue, 4/25 L10: Thur, 4/27 L11: Tues, 5/2 L12: Thur, 5/4 L13: Tues, 5/9 L14: Thur, 5/11 L15: Tue, 5/16 L16: Thur, 5/18 L17: Tues, 5/23 L18: Thur, 5/25 L19: Tue, 5/30 L20: Thur, 6/1: final presentations Class Road Map
5
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 5 Makeup Time Constraints
6
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 6 READING: Chapter 1,2 in Jordan’s book (pick up book from basement of communications copy center). Syllabus handout If you have not signed up before: –List handout: name, department, and email –List handout: regular makeup slot, and discussion section Reminder: course web page: http://ssli.ee.washington.edu/ee512 http://ssli.ee.washington.edu/ee512 TA discussions and office hours: –Office hours: Thursdays 3:30-4:30, Sieg Ground Floor Tutorial Center –Discussion Sections: Fridays 9:30-10:30, Sieg Ground Floor Tutorial Center Lecture Room Announcements
7
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 7 Bayesian Networks & CI
8
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 8 Bayesian Networks & CI
9
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 9 Bayesian Networks & CI
10
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 10 Pictorial d-separation: blocked/unblocked paths Blocked Paths Unblocked Paths
11
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 11 Three 3-node examples of BNs and their conditional independence statements. V1V1 V2V2 V3V3 V1V1 V2V2 V3V3 V1V1 V2V2 V3V3 Three Canonical Cases
12
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 12 Case 1 Markov Chain
13
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 13 Case 2 Still a Markov Chain
14
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 14 Case 3 NOT a Markov Chain
15
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 15 SUVs Greenhouse Gasses Global Warming Lung Cancer Smoking Bad Breath GeneticsCancerSmoking Examples of the three cases
16
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 16 Bayes Ball
17
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 17 Ex: What are some conditional independences? D. Poole M. Jordan
18
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 18 Two Views of a Family
19
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 19 Undirected Models
20
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 20 Markov Random Fields
21
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 21 Happy Families MRFs BNs Decomposable models
22
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 22 Can BNs represent C 4 ?
23
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 23 Markov Random Fields->Graph Theory
24
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 24 Examples Subgraph Cycle
25
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 25 Markov Random Fields
26
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 26 Markov Random Fields
27
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 27 Markov Random Fields: interaction
28
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 28 Markov Random Fields: Image Processing Derin, 1989
29
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 29 Summarizing Derin, 1989
30
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 30 Computing Probabilities
31
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 31 Computing Probabilities
32
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 32 Computing Probabilities
33
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 33 Note on sums and evidence
34
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 34 Graph Algorithm Equivalent: Elimination
35
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 35 B A C D EFG HI 1 2 3 4 7 5 6 8 9 Order the Nodes Eliminate the nodes in order Elimination Example: different graph
36
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 36 Result of Elimination
37
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 37 Elimination
38
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 38 Elimination & Directed Graphs
39
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 39 Immoral Elimination
40
Lec 2: March 30th, 2006EE512 - Graphical Models - J. BilmesPage 40 Variable Elimination Algorithm
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.