Package org.apache.wayang.ml4all.algorithms.sgd

Class WeightsUpdate

java.lang.Object

org.apache.wayang.ml4all.abstraction.api.LogicalOperator

org.apache.wayang.ml4all.abstraction.api.UpdateLocal<double[],double[]>

org.apache.wayang.ml4all.algorithms.sgd.WeightsUpdate

All Implemented Interfaces:

java.io.Serializable

public class WeightsUpdate
extends UpdateLocal<double[],double[]>

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor	Description
WeightsUpdate()
WeightsUpdate(double stepSize, double regulizer)

Method Summary

Modifier and Type	Method	Description
ML4allModel	assign(double[] input, ML4allModel model)	Assigns the new value of the global variable to the ML4allModel
double[]	process(double[] input, ML4allModel model)	Computes the new value of the global variable

Methods inherited from class org.apache.wayang.ml4all.abstraction.api.LogicalOperator

finalise, initialise

Methods inherited from class java.lang.Object

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

Constructor Detail

WeightsUpdate

public WeightsUpdate()

WeightsUpdate

public WeightsUpdate(double stepSize,
                     double regulizer)

Method Detail

process

public double[] process(double[] input,
                        ML4allModel model)

Description copied from class: UpdateLocal

Computes the new value of the global variable

Specified by:

process in class UpdateLocal<double[],double[]>

Parameters:

input - the ouput of the aggregate of the Compute

assign

public ML4allModel assign(double[] input,
                          ML4allModel model)

Description copied from class: UpdateLocal

Assigns the new value of the global variable to the ML4allModel

Specified by:

assign in class UpdateLocal<double[],double[]>

Parameters:

input - the output of the process method

Returns:

the new ML4allModel