Scribd
Upload
100 views20 pages

Causal Networks

This document discusses Bayesian approaches for learning causal networks from data. It explains that causal networks differ from acausal networks and describes circumstances where acausal network learning methods can be applied to causal networks. Specifically, it states that learning causal networks reduces to learning acausal networks when the assumptions of mechanism independence, component independence, parameter independence, and parameter modularity hold. The document provides examples of causal networks and influence diagrams and describes steps for building influence diagrams and placing them in canonical form.

Uploaded by

Vaijayanthi S
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
100 views20 pages

Causal Networks

This document discusses Bayesian approaches for learning causal networks from data. It explains that causal networks differ from acausal networks and describes circumstances where acausal network learning methods can be applied to causal networks. Specifically, it states that learning causal networks reduces to learning acausal networks when the assumptions of mechanism independence, component independence, parameter independence, and parameter modularity hold. The document provides examples of causal networks and influence diagrams and describes steps for building influence diagrams and placing them in canonical form.

Uploaded by

Vaijayanthi S
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 20

A Bayesian Approach to Learning

Causal networks

Unit 1

07/25/23 1
Objectives

• Showing that causal networks are different from a causal


ones

• Identification of circumstances in which methods for


learning acausal networks
are applicable to learning causal networks

07/25/23 2
A Causal Network is……

A directed acyclic graph where


Nodes correspond to chance variables in U
Non root node is a direct causal effect of its parents.

07/25/23 3
Causal Bayesian Networks and
Influence diagrams

A Causal Network :

b f

07/25/23 4
Some new terms :

• Unresponsiveness.
• Set decision
• Mapping variable

07/25/23 5
What is an Influence Diagram ?

A model for the domain U U D having a


structural component
probabilistic component

07/25/23 6
An Example

^b b() ^f f()

b
f

s
^s s(b,f)

^m m m(s)

07/25/23 7
Building an Influence diagram

Steps involved :
• Add a node to the diagram corresponding to each variable
in U U D
• Order the variables so that the unresponsiveness to D
comes first.
• For each Xi do
Add a causal mechanism node
Make Xi a deterministic function of Ci U Xi(Ci)where
Ci is a causal mechanism node.
Finally Assess the dependencies among the variables that are unresponsive D.

Haimonti Dutta, Department Of Computer and


07/25/23 8
Information Science
Influence diagrams in canonical forms

Conditions :
• Chance nodes descendents of D are decision nodes

• Descendents of decision nodes are deterministic nodes

Haimonti Dutta, Department Of Computer and


07/25/23 9
Information Science
Learning Influence diagrams

Observations :
Information arcs and predecessors of a utility node are not
learned

We learn only the relevance arc structure and the physical


probability

We also know the states of all the decision variables and


thus have a complete data for D in every case of the data
base.

07/25/23 10
Hence…

The problem of learning influence diagrams for the domain


UUD

reduces to

Learning acausal bayesian networks for


U UD where decision variables are interpreted as chance
variables

07/25/23 11
Learning Causal Networks

An example :
Decision to quit
smoking
x
do we get lung
cancer before
sixty?
y

07/25/23 12
The problem :

We cannot fully observe the


mapping variable y(x)

07/25/23 13
Mechanism Components

What are they?

07/25/23 14
Decomposition of the mapping
variable y(x)

x y(x=0) y(x=1)

ŷ
y

07/25/23 15
Component Independence

Assumption that the mechanism components are


independent.

y(x=0) Y(x=1)
x

ŷ
y

07/25/23 16
Another Problem

The problem : Dependent Mechanisms

A solution :Introduce additional domain variables in order


to render mechanisms independent

But….

We may not be able to observe the variables we introduce.

07/25/23 17
Learning in a causal network reduces to learning of acausal network
when
• Mechanism Independence
• Component Independence and
• Parameter Independence

07/25/23 18
Learning Causal Network structure

We can use prior network methodology to establish priors


for causal network learning provided the following holds:
• Mechanism independence
• Component independence
• Parameter independence
• Parameter modularity

07/25/23 19
Conclusion

Some important points of focus :


Mechanism Independence
Component Independence
Parameter Independence
Parameter Modularity

We use the above to learn causal networks from acausal


networks

07/25/23 20

You might also like