com.github.pmerienne.trident.ml.regression

Class PARegressor

java.lang.Object

com.github.pmerienne.trident.ml.regression.PARegressor

All Implemented Interfaces:

Regressor, Serializable

public class PARegressor
extends Object
implements Regressor

Passive-Aggresive regressor.

Author:

pmerienne

See Also:

Passive-Aggressive Algorithms Koby Crammer, Ofer Dekel, Joseph Keshet, Shai Shalev-Shwartz, Yoram Singer; 7(Mar):551--585, 2006., Serialized Form

Constructor Summary

Constructors

Constructor and Description
PARegressor()
PARegressor(Double epsilon)

Method Summary

Methods

Modifier and Type	Method and Description
Double	getEpsilon()
double[]	getWeights()
protected void	init(int featureSize)
Double	predict(double[] features)
void	reset()
void	setEpsilon(Double epsilon)
void	setWeights(double[] weights)
String	toString()
void	update(Double expected, double[] features)

Methods inherited from class java.lang.Object

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

Constructor Detail

PARegressor

public PARegressor()

PARegressor

public PARegressor(Double epsilon)

Method Detail

predict

public Double predict(double[] features)

Specified by:

predict in interface Regressor

update

public void update(Double expected,
          double[] features)

Specified by:

update in interface Regressor

init

protected void init(int featureSize)

reset

public void reset()

Specified by:

reset in interface Regressor

getWeights

public double[] getWeights()

setWeights

public void setWeights(double[] weights)

getEpsilon

public Double getEpsilon()

setEpsilon

public void setEpsilon(Double epsilon)

toString

public String toString()

Overrides:

toString in class Object