org.knime.base.node.mine.decisiontree2.learner
Class SplitFinder

java.lang.Object
  extended by org.knime.base.node.mine.decisiontree2.learner.SplitFinder

public class SplitFinder
extends Object

Finds the best split for a given InMemoryTable. The results can be retrieved via getter methdods.

Author:
Christoph Sieb, University of Konstanz

Constructor Summary
SplitFinder(InMemoryTable table, SplitQualityMeasure splitQualityMeasure, boolean averageSplitpoint, double minObjectsCount, boolean binaryNominalSplits, int maxNumNominalsForCompleteComputation)
          Finds the best split for the given data.
 
Method Summary
 Split getSplit()
          Returns the split evaluated as the best for the given data.
 int getSplitAttributeIndex()
          Returns the attribute list that determines the split.
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

SplitFinder

public SplitFinder(InMemoryTable table,
                   SplitQualityMeasure splitQualityMeasure,
                   boolean averageSplitpoint,
                   double minObjectsCount,
                   boolean binaryNominalSplits,
                   int maxNumNominalsForCompleteComputation)
Finds the best split for the given data.

Parameters:
table - the data table for which to find the best split attribute and for this the best split
splitQualityMeasure - the quality measure (e.g. gini or gain ratio)
averageSplitpoint - if true, the split point is set as the average of the partition borders, else the upper value of the lower partition is used
minObjectsCount - minimum number of examples for a partition
binaryNominalSplits - if true, nominal attributes are split according to binary subsets, else each nominal value represents one branch
maxNumNominalsForCompleteComputation - the maximum number of nominal values for which all subsets are calculated (results in the optimal binary split); this parameter is only use if binaryNominalSplits is true; if the number of nominal values is higher, a heuristic is applied
Method Detail

getSplit

public Split getSplit()
Returns the split evaluated as the best for the given data.

Returns:
the split evaluated as the best for the given data

getSplitAttributeIndex

public int getSplitAttributeIndex()
Returns the attribute list that determines the split.

Returns:
the attribute list that determines the split


Copyright, 2003 - 2010. All rights reserved.
University of Konstanz, Germany.
Chair for Bioinformatics and Information Mining, Prof. Dr. Michael R. Berthold.
You may not modify, publish, transmit, transfer or sell, reproduce, create derivative works from, distribute, perform, display, or in any way exploit any of the content, in whole or in part, except as otherwise expressly permitted in writing by the copyright owner or as specified in the license file distributed with this product.