All Packages  Class Hierarchy  This Package  Previous  Next  Index  WEKA's home

Class weka.classifiers.AdditiveRegression

java.lang.Object
    |
    +----weka.classifiers.Classifier
            |
            +----weka.classifiers.AdditiveRegression

public class AdditiveRegression
extends Classifier
implements OptionHandler, AdditionalMeasureProducer
Meta classifier that enhances the performance of a regression base classifier. Each iteration fits a model to the residuals left by the classifier on the previous iteration. Prediction is accomplished by adding the predictions of each classifier. Smoothing is accomplished through varying the shrinkage (learning rate) parameter.

 Analysing:  Root_relative_squared_error
 Datasets:   36
 Resultsets: 2
 Confidence: 0.05 (two tailed)
 Date:       10/13/00 10:00 AM


 Dataset                   (1) m5.M5Prim | (2) AdditiveRegression -S 0.7 \
                                         |    -B weka.classifiers.m5.M5Prime 
                          ----------------------------
 auto93.names              (10)    54.4  |    49.41 * 
 autoHorse.names           (10)    32.76 |    26.34 * 
 autoMpg.names             (10)    35.32 |    34.84 * 
 autoPrice.names           (10)    40.01 |    36.57 * 
 baskball                  (10)    79.46 |    79.85   
 bodyfat.names             (10)    10.38 |    11.41 v 
 bolts                     (10)    19.29 |    12.61 * 
 breastTumor               (10)    96.95 |    96.23 * 
 cholesterol               (10)   101.03 |    98.88 * 
 cleveland                 (10)    71.29 |    70.87 * 
 cloud                     (10)    38.82 |    39.18   
 cpu                       (10)    22.26 |    14.74 * 
 detroit                   (10)   228.16 |    83.7  * 
 echoMonths                (10)    71.52 |    69.15 * 
 elusage                   (10)    48.94 |    49.03   
 fishcatch                 (10)    16.61 |    15.36 * 
 fruitfly                  (10)   100    |   100    * 
 gascons                   (10)    18.72 |    14.26 * 
 housing                   (10)    38.62 |    36.53 * 
 hungarian                 (10)    74.67 |    72.19 * 
 longley                   (10)    31.23 |    28.26 * 
 lowbwt                    (10)    62.26 |    61.48 * 
 mbagrade                  (10)    89.2  |    89.2    
 meta                      (10)   163.15 |   188.28 v 
 pbc                       (10)    81.35 |    79.4  * 
 pharynx                   (10)   105.41 |   105.03   
 pollution                 (10)    72.24 |    68.16 * 
 pwLinear                  (10)    32.42 |    33.33 v 
 quake                     (10)   100.21 |    99.93   
 schlvote                  (10)    92.41 |    98.23 v 
 sensory                   (10)    88.03 |    87.94   
 servo                     (10)    37.07 |    35.5  * 
 sleep                     (10)    70.17 |    71.65   
 strike                    (10)    84.98 |    83.96 * 
 veteran                   (10)    90.61 |    88.77 * 
 vineyard                  (10)    79.41 |    73.95 * 
                        ----------------------------
                              (v| |*) |   (4|8|24) 

 

For more information see:

Friedman, J.H. (1999). Stochastic Gradient Boosting. Technical Report Stanford University. http://www-stat.stanford.edu/~jhf/ftp/stobst.ps.

Valid options from the command line are:

-B classifierstring
Classifierstring should contain the full class name of a classifier followed by options to the classifier. (required).

-S shrinkage rate
Smaller values help prevent overfitting and have a smoothing effect (but increase learning time). (default = 1.0, ie no shrinkage).

-M max models
Set the maximum number of models to generate. Values <= 0 indicate no maximum, ie keep going until the reduction in error threshold is reached. (default = -1).

-D
Debugging output.

Version:
$Revision: 1.6 $
Author:
Mark Hall (mhall@cs.waikato.ac.nz)

Constructor Index

 o AdditiveRegression()
Default constructor specifying DecisionStump as the classifier
 o AdditiveRegression(Classifier)
Constructor which takes base classifier as argument.

Method Index

 o buildClassifier(Instances)
Build the classifier on the supplied data
 o classifierTipText()
Returns the tip text for this property
 o classifyInstance(Instance)
Classify an instance.
 o debugTipText()
Returns the tip text for this property
 o enumerateMeasures()
Returns an enumeration of the additional measure names
 o getClassifier()
Gets the classifier used.
 o getDebug()
Gets whether debugging has been turned on
 o getMaxModels()
Get the max number of models to generate
 o getMeasure(String)
Returns the value of the named measure
 o getOptions()
Gets the current settings of the Classifier.
 o getShrinkage()
Get the shrinkage rate.
 o globalInfo()
Returns a string describing this attribute evaluator
 o listOptions()
Returns an enumeration describing the available options
 o main(String[])
Main method for testing this class.
 o maxModelsTipText()
Returns the tip text for this property
 o measureNumIterations()
return the number of iterations (base classifiers) completed
 o setClassifier(Classifier)
Sets the classifier
 o setDebug(boolean)
Set whether debugging output is produced.
 o setMaxModels(int)
Set the maximum number of models to generate
 o setOptions(String[])
Parses a given list of options.
 o setShrinkage(double)
Set the shrinkage parameter
 o shrinkageTipText()
Returns the tip text for this property
 o toString()
Returns textual description of the classifier.

Constructor Detail

 o AdditiveRegression
public AdditiveRegression()
          Default constructor specifying DecisionStump as the classifier
 o AdditiveRegression
public AdditiveRegression(Classifier classifier)
          Constructor which takes base classifier as argument.
Parameters:
classifier - the base classifier to use

Method Detail

 o globalInfo
public java.lang.String globalInfo()
          Returns a string describing this attribute evaluator
Returns:
a description of the evaluator suitable for displaying in the explorer/experimenter gui
 o listOptions
public java.util.Enumeration listOptions()
          Returns an enumeration describing the available options
Returns:
an enumeration of all the available options
 o setOptions
public void setOptions(java.lang.String options[]) throws java.lang.Exception
          Parses a given list of options. Valid options are:

-B classifierstring
Classifierstring should contain the full class name of a classifier followed by options to the classifier. (required).

-S shrinkage rate
Smaller values help prevent overfitting and have a smoothing effect (but increase learning time). (default = 1.0, ie. no shrinkage).

-D
Debugging output.

-M max models
Set the maximum number of models to generate. Values <= 0 indicate no maximum, ie keep going until the reduction in error threshold is reached. (default = -1).

Parameters:
options - the list of options as an array of strings
Throws:
java.lang.Exception - if an option is not supported
 o getOptions
public java.lang.String[] getOptions()
          Gets the current settings of the Classifier.
Returns:
an array of strings suitable for passing to setOptions
 o debugTipText
public java.lang.String debugTipText()
          Returns the tip text for this property
Returns:
tip text for this property suitable for displaying in the explorer/experimenter gui
 o setDebug
public void setDebug(boolean d)
          Set whether debugging output is produced.
Parameters:
d - true if debugging output is to be produced
 o getDebug
public boolean getDebug()
          Gets whether debugging has been turned on
Returns:
true if debugging has been turned on
 o classifierTipText
public java.lang.String classifierTipText()
          Returns the tip text for this property
Returns:
tip text for this property suitable for displaying in the explorer/experimenter gui
 o setClassifier
public void setClassifier(Classifier classifier)
          Sets the classifier
Parameters:
classifier - the classifier with all options set.
 o getClassifier
public Classifier getClassifier()
          Gets the classifier used.
Returns:
the classifier
 o maxModelsTipText
public java.lang.String maxModelsTipText()
          Returns the tip text for this property
Returns:
tip text for this property suitable for displaying in the explorer/experimenter gui
 o setMaxModels
public void setMaxModels(int maxM)
          Set the maximum number of models to generate
Parameters:
maxM - the maximum number of models
 o getMaxModels
public int getMaxModels()
          Get the max number of models to generate
Returns:
the max number of models to generate
 o shrinkageTipText
public java.lang.String shrinkageTipText()
          Returns the tip text for this property
Returns:
tip text for this property suitable for displaying in the explorer/experimenter gui
 o setShrinkage
public void setShrinkage(double l)
          Set the shrinkage parameter
Parameters:
l - the shrinkage rate.
 o getShrinkage
public double getShrinkage()
          Get the shrinkage rate.
Returns:
the value of the learning rate
 o buildClassifier
public void buildClassifier(Instances data) throws java.lang.Exception
          Build the classifier on the supplied data
Parameters:
data - the training data
Throws:
java.lang.Exception - if the classifier could not be built successfully
Overrides:
buildClassifier in class Classifier
 o classifyInstance
public double classifyInstance(Instance inst) throws java.lang.Exception
          Classify an instance.
Parameters:
inst - the instance to predict
Returns:
a prediction for the instance
Throws:
java.lang.Exception - if an error occurs
Overrides:
classifyInstance in class Classifier
 o enumerateMeasures
public java.util.Enumeration enumerateMeasures()
          Returns an enumeration of the additional measure names
Returns:
an enumeration of the measure names
 o getMeasure
public double getMeasure(java.lang.String additionalMeasureName)
          Returns the value of the named measure
Parameters:
measureName - the name of the measure to query for its value
Returns:
the value of the named measure
Throws:
java.lang.IllegalArgumentException - if the named measure is not supported
 o measureNumIterations
public double measureNumIterations()
          return the number of iterations (base classifiers) completed
Returns:
the number of iterations (same as number of base classifier models)
 o toString
public java.lang.String toString()
          Returns textual description of the classifier.
Returns:
a description of the classifier as a string
Overrides:
toString in class java.lang.Object
 o main
public static void main(java.lang.String argv[])
          Main method for testing this class.
Parameters:
argv - should contain the following arguments: -t training file [-T test file] [-c class index]

All Packages  Class Hierarchy  This Package  Previous  Next  Index  WEKA's home