30
1 On Distributing a On Distributing a Bayesian Network Bayesian Network Thor Whalen Metron, Inc.

On Distributing a Bayesian Network

  • Upload
    lonato

  • View
    24

  • Download
    0

Embed Size (px)

DESCRIPTION

On Distributing a Bayesian Network. Thor Whalen Metron, Inc. Major discussion points. Functionality afforded by old code Issues and problems with old approach Functionality trying to add Functionality successfully added What is not yet working correctly What sill have to try to add. - PowerPoint PPT Presentation

Citation preview

Page 1: On Distributing a Bayesian Network

1

On Distributing a Bayesian On Distributing a Bayesian NetworkNetwork

Thor Whalen

Metron, Inc.

Page 2: On Distributing a Bayesian Network

2

Major discussion points

• Functionality afforded by old code

• Issues and problems with old approach

• Functionality trying to add

• Functionality successfully added

• What is not yet working correctly

• What sill have to try to add

Page 3: On Distributing a Bayesian Network

3

Outline

• Need a Review of JT?

• Distributing a BN using the JT: Issues.

• Directions for Solutions

• Matlab: New functionality and experiments

• And now what?

Page 4: On Distributing a Bayesian Network

4

a

b c

d e

f

g

h

Secondary Structure/Junction Tree• multi-dim. random variables• joint probabilities (potentials)

Bayesian Network• one-dim. random variables• conditional probabilities

abd

ade

ace

ceg

eghdef

ad ae ce

de eg

Page 5: On Distributing a Bayesian Network

5

Building a Junction Tree

DAG

Moral Graph

Triangulated Graph

Junction Tree

Identifying Cliques

Page 6: On Distributing a Bayesian Network

6

Step 1: Moralization

a

b c

d e

f

g

h

a

b c

d e

f

g

h

a

b c

d e

f

g

h

1. For all w V:• For all u,vpa(w) add an edge e=u-v.

2. Undirect all edges.

GMG = ( V , E )

Page 7: On Distributing a Bayesian Network

7

Step 2: Triangulation

Add edges to GM such that there is no cyclewith length 4 that does not contain a chord.

NO YES

a

b c

d e

f

g

h

a

b c

d e

f

g

h

GM GT

Page 8: On Distributing a Bayesian Network

8

a

b

d

a

c

e

d e

f

a

d e

e

g

h

c

e

g

a

b c

d e

f

g

h

Bayesian NetworkG = ( V , E )

a

b c

d e

f

g

h

a

b c

d e

f

g

h

Moral graph GM Triangulated graph GT

abd

ade

ace

ceg

eghdef

ad ae ce

de eg

seperators

Junction graph GJ (not complete)e.g. ceg egh = eg Cliques

e

e

e

a

e

Page 9: On Distributing a Bayesian Network

9

Junction graph GJ (not complete)

abd

ade

ace

ceg

eghdef

ad ae ce

de eg

e

e

e

a

e

abd

ade

ace

ceg

eghdef

ad ae ce

de eg

Junction tree GJT

There are several methods to find MST.

Kruskal’s algorithm: choose successively a link of

maximal weight unless it creates a cycle.

Page 10: On Distributing a Bayesian Network

10

a

b

d

a

c

e

d e

f

a

d e

e

g

h

c

e

g

abd

ade

ace

ceg

eghdef

ad ae ce

de eg

sepsets

In JT cliquesbecomesvertices

GJT

Ex: ceg egh = eg

Page 11: On Distributing a Bayesian Network

11

Message Passing from clique A to clique B

1. Project the potential of A into SAB

2. Absorb the potential of SAB into B

Projection Absorption

Propagating potentialsPropagating potentials

Page 12: On Distributing a Bayesian Network

12

1. COLLECT-EVIDENCE messages 1-52. DISTRIBUTE-EVIDENCE messages 6-10

Global PropagationGlobal Propagation

32

5

1 48 10

9

7

6

Root

abd

ade

ace

ceg

eghdef

ad ae ce

de eg

Page 13: On Distributing a Bayesian Network

13

Using the JT for the MSBN

• How?

• Issue: Clique Granularity

• Issue: Clique Association

• Issue: MSBN must have tree structure

Page 14: On Distributing a Bayesian Network

14

Take a JT

Page 15: On Distributing a Bayesian Network

15

Partition JT into sub-trees

Page 16: On Distributing a Bayesian Network

16

Issue: Clique Granularity

Page 17: On Distributing a Bayesian Network

17

Issue: Clique Granularity

Page 18: On Distributing a Bayesian Network

18

Issue: Clique association

Page 19: On Distributing a Bayesian Network

19

Issue: Clique association

Page 20: On Distributing a Bayesian Network

20

Issue: MSBN has a tree structure

Page 21: On Distributing a Bayesian Network

21

Issue: MSBN has a tree structure

Can’t communicate here!

Page 22: On Distributing a Bayesian Network

22

Issue: MSBN has a tree structure

Can’t communicate here!

Though these two subnets may share many variables.

Page 23: On Distributing a Bayesian Network

23

Directions for Solutions

• Controlling Granularity

• Increasing Granularity

• Choosing the JT

• Alternative communication graphs

Page 24: On Distributing a Bayesian Network

24

Controlling Granularity

• Moralization and Triangulation → Cliques

• Moralization: No choice.

• Triangulation: Some choice.

• Triangulate keeping the subnets in mind.

Page 25: On Distributing a Bayesian Network

25

Increasing Granularity

• Can we break down cliques?

Conjecture:Given two sets of random variables X and Y, let

dist(P(X-Y|Y),P(X-Y|X Y))(X,Y) =

vol(X Y)

Given a clique Z whose set of random variables Z is covered by sets X and Y. If ∆(X,Y) is small enough then we may replace Z by X and Y.

Z X Y

Page 26: On Distributing a Bayesian Network

26

Choosing The JT

• Once the cliques have been formed (hence the JG), any maximal weight spanning tree of the JG will do for the JT.

• Again, we should choose this tree so as to reduce the overhead when connecting subnets internally.

Page 27: On Distributing a Bayesian Network

27

Alternative communication Graphs

We raise the question as to wether it is possible to perform intra- and extra-subnet calibration using some non-tree sub-graph of the JG…

Page 28: On Distributing a Bayesian Network

28

MatLab: The new Stuff:

• Netica can talk to the Matlab tool.

• View/Interact tool is nicer

• JG for communication graph… (or not)

• Query cliques

Page 29: On Distributing a Bayesian Network

29

Propagating with cycles

ABC BCE

BDE DEF

BC

EBE

B

DE

Page 30: On Distributing a Bayesian Network

30

Mysterious convergence

ABC CE

BDE DEF

C

EE

B

DE