Module es.upm.fi.cig.multictbnc

Package es.upm.fi.cig.multictbnc.learning.structure.constraints.CTBNC

Class MaxKCTBNC

java.lang.Object

es.upm.fi.cig.multictbnc.learning.structure.constraints.AbstractStructureConstraints

es.upm.fi.cig.multictbnc.learning.structure.constraints.CTBNC.MaxKCTBNC

All Implemented Interfaces:

StructureConstraints

public class MaxKCTBNC
extends AbstractStructureConstraints

Implements the structure constraints of a Max-k continuous-time Bayesian network classifier, i.e., a CTBNC where the number of parents of the feature nodes is bounded by a positive number (Codecasa and Stella, 2014).

Constructor Summary

Constructors

Constructor	Description
MaxKCTBNC(int maxK)	Receives the maximum number of parents the nodes can have.

Method Summary

Modifier and Type	Method	Description
boolean	isStructureLegal(boolean[][] adjacencyMatrix, NodeIndexer<? extends Node> nodeIndexer)	Determines if the structure of a PGM is legal.

Methods inherited from class es.upm.fi.cig.multictbnc.learning.structure.constraints.AbstractStructureConstraints

initialiseStructure, uniqueStructure

Methods inherited from class java.lang.Object

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

Constructor Details

MaxKCTBNC

public MaxKCTBNC(int maxK)

Receives the maximum number of parents the nodes can have. If a positive number is not given, one parent is used.

Parameters:

maxK - maximum number of parents the nodes of feature variables can have (without including nodes of class variables)

Method Details

isStructureLegal

public boolean isStructureLegal(boolean[][] adjacencyMatrix,
 NodeIndexer<? extends Node> nodeIndexer)

Description copied from interface: StructureConstraints

Determines if the structure of a PGM is legal.

Parameters:

adjacencyMatrix - adjacency matrix

nodeIndexer - node indexer that allows access to information about the nodes

Returns:

true if the structure is legal, false otherwise