keel.Algorithms.Decision_Trees.SLIQ

Class SLIQ

java.lang.Object

keel.Algorithms.Decision_Trees.SLIQ.Algorithm

keel.Algorithms.Decision_Trees.SLIQ.SLIQ

public class SLIQ
extends Algorithm

Implementation of the SLIQ algorithm.

Version:

1.0 (28/12/09 - 10/1/10)

Author:

Francisco Charte Ojeda (UJA)

Field Summary

Fields inherited from class keel.Algorithms.Decision_Trees.SLIQ.Algorithm

correct, log, modelDataset, modelFileName, resultFileName, startTime, testCorrect, testDataset, testFileName, testOutputFileName, trainDataset, trainFileName, trainOutputFileName

Constructor Summary

Constructors

Constructor and Description
SLIQ(java.lang.String paramFile) Constructor.

Method Summary

Modifier and Type	Method and Description
protected void	agregaDatos(Node padre, Node hijo) Aggregates the data of the given child that will be pruned into the father given.
protected void	aplicaMejorCorte(Node nodo) Método que aplica en un nodo el mejor corte obtenido previamente Apply the best cut fouded to divide the given node.
protected void	aplicaMejorCorteContinuo(Node nodo) Divides the node in the more optimal way possible (Continuous attribute).
protected void	aplicaMejorCorteDiscreto(Node nodo) Divides the node in the more optimal way possible (Discrete attribute).
protected void	calculaMejorCorte(int indAtributo, Node nodo) Computes the best cut of the given attribute for the given node.
void	cuentaNodosHojas(Node node) Counts the number of nodes pending from the given node.
int	evaluateItemset(Itemset itemset, Node node) Returns the predicted class of the given example obtained from the given node.
protected void	generaListas() Constructs the lists of attriibutes and classes used during SLIQ execution.
void	generateTree() Generates the tree with the SLIQ algorithm.
static void	main(java.lang.String[] args) Main function.
protected void	podaArbol() Prunes the tree.
protected void	podaNodoCompleto(Node padre) Prunes the node given as parameter, deleting its children.
protected void	podaNodoParcial(Node padre, int indHijo) Prunes the node given as parameter, deleting the given child.
void	printResult() Writes the statistical measurements obtained on the output file.
void	printTest() Evaluates the test dataset and writes the results on the output file.
void	printTrain() Evaluates the training dataset and writes the results on the output file.
protected void	setOptions(java.io.StreamTokenizer options) Reads the parameters used by the algorith.

Methods inherited from class keel.Algorithms.Decision_Trees.SLIQ.Algorithm

getHeader, initTokenizer

Methods inherited from class java.lang.Object

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

Constructor Detail

SLIQ

public SLIQ(java.lang.String paramFile)

Constructor.

Parameters:

paramFile - parameters file.

Method Detail

generaListas

protected void generaListas()

Constructs the lists of attriibutes and classes used during SLIQ execution.

generateTree

public void generateTree()

Generates the tree with the SLIQ algorithm.

podaArbol

protected void podaArbol()

Prunes the tree.

podaNodoCompleto

protected void podaNodoCompleto(Node padre)

Prunes the node given as parameter, deleting its children.

Parameters:

padre - given node to be pruned.

podaNodoParcial

protected void podaNodoParcial(Node padre,
                               int indHijo)

Prunes the node given as parameter, deleting the given child.

Parameters:

padre - given father to be pruned.

indHijo - given child index to delete.

agregaDatos

protected void agregaDatos(Node padre,
                           Node hijo)

Aggregates the data of the given child that will be pruned into the father given.

Parameters:

padre - Father node.

hijo - Child node.

calculaMejorCorte

protected void calculaMejorCorte(int indAtributo,
                                 Node nodo)

Computes the best cut of the given attribute for the given node.

Parameters:

indAtributo - Attribute index.

nodo - Node to be divided.

aplicaMejorCorte

protected void aplicaMejorCorte(Node nodo)

Método que aplica en un nodo el mejor corte obtenido previamente Apply the best cut fouded to divide the given node.

Parameters:

nodo - node to be divided.

aplicaMejorCorteDiscreto

protected void aplicaMejorCorteDiscreto(Node nodo)

Divides the node in the more optimal way possible (Discrete attribute).

Parameters:

nodo - node to be divided.

aplicaMejorCorteContinuo

protected void aplicaMejorCorteContinuo(Node nodo)

Divides the node in the more optimal way possible (Continuous attribute).

Parameters:

nodo - node to be divided.

evaluateItemset

public int evaluateItemset(Itemset itemset,
                           Node node)

Returns the predicted class of the given example obtained from the given node.

Parameters:

itemset - given example.

node - given node.

Returns:

predicted class.

cuentaNodosHojas

public void cuentaNodosHojas(Node node)

Counts the number of nodes pending from the given node.

Parameters:

node - given node.

printResult

public void printResult()
                 throws java.io.IOException

Writes the statistical measurements obtained on the output file.

Specified by:

printResult in class Algorithm

Throws:

java.io.IOException - if the file can not be written.

printTrain

public void printTrain()

Evaluates the training dataset and writes the results on the output file.

Specified by:

printTrain in class Algorithm

printTest

public void printTest()

Evaluates the test dataset and writes the results on the output file.

Specified by:

printTest in class Algorithm

setOptions

protected void setOptions(java.io.StreamTokenizer options)
                   throws java.lang.Exception

Reads the parameters used by the algorith.

Specified by:

setOptions in class Algorithm

Parameters:

options - StreamTokenizer used to read the different parameters

Throws:

java.lang.Exception - if the input file format is wrong.

main

public static void main(java.lang.String[] args)

Main function. Executes the SLIQ algorithm with the configuration file given.

Parameters:

args - Arguments of the program (a configuration script, generally)