weka.classifiers.functions

Class MLPRegressor

java.lang.Object

weka.classifiers.AbstractClassifier

weka.classifiers.RandomizableClassifier

weka.classifiers.functions.MLPModel

weka.classifiers.functions.MLPRegressor

All Implemented Interfaces:

java.io.Serializable, java.lang.Cloneable, weka.classifiers.Classifier, weka.core.BatchPredictor, weka.core.CapabilitiesHandler, weka.core.CapabilitiesIgnorer, weka.core.CommandlineRunnable, weka.core.OptionHandler, weka.core.Randomizable, weka.core.RevisionHandler, weka.core.WeightedInstancesHandler

public class MLPRegressor
extends MLPModel
implements weka.core.WeightedInstancesHandler

Trains a multilayer perceptron with one hidden layer using WEKA's Optimization class by minimizing the given loss function plus a quadratic penalty with the BFGS method. Note that all attributes are standardized, including the target. There are several parameters. The ridge parameter is used to determine the penalty on the size of the weights. The number of hidden units can also be specified. Note that large numbers produce long training times. Finally, it is possible to use conjugate gradient descent rather than BFGS updates, which may be faster for cases with many parameters. To improve speed, an approximate version of the logistic function is used as the default activation function for the hidden layer, but other activation functions can be specified. In the output layer, the sigmoid function is used for classification. If the approximate sigmoid is specified for the hidden layers, it is also used for the output layer. For regression, the identity function is used activation function in the output layer. Also, if delta values in the backpropagation step are within the user-specified tolerance, the gradient is not updated for that particular instance, which saves some additional time. Parallel calculation of loss function and gradient is possible when multiple CPU cores are present. Data is split into batches and processed in separate threads in this case. Note that this only improves runtime for larger datasets. Nominal attributes are processed using the unsupervised NominalToBinary filter and missing values are replaced globally using ReplaceMissingValues.

Valid options are:

 -N <int>
  Number of hidden units (default is 2).

 -R <double>
  Ridge factor for quadratic penalty on weights (default is 0.01).

 -O <double>
  Tolerance parameter for delta values (default is 1.0e-6).

 -G
  Use conjugate gradient descent (recommended for many attributes).

 -P <int>
  The size of the thread pool, for example, the number of cores in the CPU. (default 1)

 -E <int>
  The number of threads to use, which should be >= size of thread pool. (default 1)

 -L <classname and parameters>
  The loss function to use.
  (default: weka.classifiers.functions.loss.SquaredError)

 -A <classname and parameters>
  The activation function to use.
  (default: weka.classifiers.functions.activation.ApproximateSigmoid)

 -S <num>
  Random number seed.
  (default 1)

 -output-debug-info
  If set, classifier is run in debug mode and
  may output additional info to the console

 -do-not-check-capabilities
  If set, classifier capabilities are not checked before classifier is built
  (use with caution).

 -num-decimal-places
  The number of decimal places for the output of numbers in the model (default 2).

Version:

$Revision: 12069 $

Author:

Eibe Frank (eibe@cs.waikato.ac.nz)

See Also:

Serialized Form

Field Summary

Fields inherited from class weka.classifiers.AbstractClassifier

BATCH_SIZE_DEFAULT, NUM_DECIMAL_PLACES_DEFAULT

Constructor Summary

Constructors

Constructor and Description
MLPRegressor()

Method Summary

Modifier and Type	Method and Description
weka.core.Capabilities	getCapabilities() Returns default capabilities of the classifier.
static void	main(java.lang.String[] argv) Main method to run the code from the command-line using the standard WEKA options.
java.lang.String	modelType() Returns the model type as a string.

Methods inherited from class weka.classifiers.functions.MLPModel

activationFunctionTipText, buildClassifier, distributionForInstance, getActivationFunction, getLossFunction, getNumFunctions, getNumThreads, getOptions, getPoolSize, getRidge, getTolerance, getUseCGD, globalInfo, listOptions, lossFunctionTipText, numFunctionsTipText, numThreadsTipText, poolSizeTipText, ridgeTipText, setActivationFunction, setLossFunction, setNumFunctions, setNumThreads, setOptions, setPoolSize, setRidge, setTolerance, setUseCGD, toleranceTipText, toString, useCGDTipText

Methods inherited from class weka.classifiers.RandomizableClassifier

getSeed, seedTipText, setSeed

Methods inherited from class weka.classifiers.AbstractClassifier

batchSizeTipText, classifyInstance, debugTipText, distributionsForInstances, doNotCheckCapabilitiesTipText, forName, getBatchSize, getDebug, getDoNotCheckCapabilities, getNumDecimalPlaces, getRevision, implementsMoreEfficientBatchPrediction, makeCopies, makeCopy, numDecimalPlacesTipText, postExecution, preExecution, run, runClassifier, setBatchSize, setDebug, setDoNotCheckCapabilities, setNumDecimalPlaces

Methods inherited from class java.lang.Object

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

Methods inherited from interface weka.core.OptionHandler

makeCopy

Constructor Detail

MLPRegressor

public MLPRegressor()

Method Detail

getCapabilities

public weka.core.Capabilities getCapabilities()

Returns default capabilities of the classifier.

Specified by:

getCapabilities in interface weka.classifiers.Classifier

Specified by:

getCapabilities in interface weka.core.CapabilitiesHandler

Overrides:

getCapabilities in class MLPModel

Returns:

the capabilities of this classifier

modelType

public java.lang.String modelType()

Returns the model type as a string.

Specified by:

modelType in class MLPModel

main

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

Main method to run the code from the command-line using the standard WEKA options.