jsat.classifiers.boosting

Class LogitBoostPL

java.lang.Object

jsat.classifiers.boosting.LogitBoost

jsat.classifiers.boosting.LogitBoostPL

All Implemented Interfaces:

Serializable, Cloneable, Classifier, Parameterized

public class LogitBoostPL
extends LogitBoost

An extension to the original LogitBoost algorithm for parallel training. This comes at an increase in classification time.
Note: LogitBoost is a semi unstable algorithm, and this method does in fact increase the instability. In most cases, similar classification results will be obtained, however - under some circumstances the performance may be significantly degraded. Especially if there is insufficient data to distribute for parallel computation. The results for LogitBoost seem to be over stated in the original paper.
See: Scalable and Parallel Boosting with MapReduce, Indranil Palit and Chandan K. Reddy, IEEE Transactions on Knowledge and Data Engineering

Author:

Edward Raff

See Also:

Serialized Form

Field Summary

Fields inherited from class jsat.classifiers.boosting.LogitBoost

baseLearner, baseLearners, fScaleConstant

Constructor Summary

Constructors

Constructor and Description
LogitBoostPL(int M)
LogitBoostPL(Regressor baseLearner, int M)

Method Summary

Modifier and Type	Method and Description
LogitBoostPL	clone()
void	trainC(ClassificationDataSet dataSet, ExecutorService threadPool) Trains the classifier and constructs a model for classification using the given data set.

Methods inherited from class jsat.classifiers.boosting.LogitBoost

classify, getMaxIterations, getModels, getParameter, getParameters, getzMax, P, setMaxIterations, setzMax, supportsWeightedData, trainC

Methods inherited from class java.lang.Object

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

Constructor Detail

LogitBoostPL

public LogitBoostPL(Regressor baseLearner,
                    int M)

LogitBoostPL

public LogitBoostPL(int M)

Method Detail

trainC

public void trainC(ClassificationDataSet dataSet,
                   ExecutorService threadPool)

Description copied from interface: Classifier

Trains the classifier and constructs a model for classification using the given data set. If the training method knows how, it will used the threadPool to conduct training in parallel. This method will block until the training has completed.

Specified by:

trainC in interface Classifier

Overrides:

trainC in class LogitBoost

Parameters:

dataSet - the data set to train on

threadPool - the source of threads to use.

clone

public LogitBoostPL clone()

Specified by:

clone in interface Classifier

Overrides:

clone in class LogitBoost