bayesnet.jayes.transformation

Class LatentDeterministicDecomposition

java.lang.Object

bayesnet.jayes.transformation.AbstractDecomposition

bayesnet.jayes.transformation.LatentDeterministicDecomposition

All Implemented Interfaces:

IDecompositionStrategy

public class LatentDeterministicDecomposition
extends AbstractDecomposition

To be used when a conditional probability density consists of mostly the same few distributions, which is often the case when a deterministic distribution is smoothed.

 Example:
  {
    { 0.99 0.01 }
    { 0.01 0.99 }
    { 0.01 0.99 }
  }
 
  would decompose to the new basis
  
  {
    { 0.01 0.99 }
    { 0.99 0.01 }
  }
  
  and 
  {
    { 0 1 }
    { 1 0 }
    { 1 0 }
  }
 

The decomposition is still applicable when there are some parts of the distribution that can not be directly expressed by one of the basis vectors. In that case, that particular distribution will be represented as non-negative linear combination of the basis vectors. Note that because of that, it is possible that the decomposition fails.

Constructor Summary

Constructors

Constructor and Description
LatentDeterministicDecomposition()

Method Summary

Methods inherited from class bayesnet.jayes.transformation.AbstractDecomposition

decompose

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

LatentDeterministicDecomposition

public LatentDeterministicDecomposition()