Package de.jungblut.math.activation

Class AbstractActivationFunction

java.lang.Object

de.jungblut.math.activation.AbstractActivationFunction

All Implemented Interfaces:

ActivationFunction

Direct Known Subclasses:

ElliotActivationFunction, LinearActivationFunction, LogActivationFunction, ReluActivationFunction, SigmoidActivationFunction, SoftMaxActivationFunction, SoftplusReluActivationFunction, StepActivationFunction, TanhActivationFunction

public abstract class AbstractActivationFunction
extends java.lang.Object
implements ActivationFunction

Implements the boiler plate code for applying functions on container classes like vectors and matrices by applying the function on every element. This implementation is aware of the type of the vector and matrix, so it is also optimized for sparse as well as dense types.

Author:

thomas.jungblut

Constructor Summary

Constructors

Constructor	Description
AbstractActivationFunction()

Method Summary

Modifier and Type	Method	Description
de.jungblut.math.DoubleMatrix	apply(de.jungblut.math.DoubleMatrix matrix)	Applies the gradient of the activation function on each element in the given matrix.
de.jungblut.math.DoubleVector	apply(de.jungblut.math.DoubleVector vector)	Applies the activation function on each element in the given vector.
de.jungblut.math.DoubleMatrix	gradient(de.jungblut.math.DoubleMatrix matrix)	Applies the gradient of the activation function on each element in the given matrix.
de.jungblut.math.DoubleVector	gradient(de.jungblut.math.DoubleVector vector)	Applies the gradient of the activation function on each element in the given vector.
protected de.jungblut.math.DoubleMatrix	newInstance(de.jungblut.math.DoubleMatrix mat)
protected de.jungblut.math.DoubleVector	newInstance(de.jungblut.math.DoubleVector v)
java.lang.String	toString()

Methods inherited from class java.lang.Object

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

Methods inherited from interface de.jungblut.math.activation.ActivationFunction

apply, gradient

Constructor Detail

AbstractActivationFunction

public AbstractActivationFunction()

Method Detail

apply

public de.jungblut.math.DoubleVector apply(de.jungblut.math.DoubleVector vector)

Description copied from interface: ActivationFunction

Applies the activation function on each element in the given vector.

Specified by:

apply in interface ActivationFunction

Parameters:

vector - the vector to apply this function on.

Returns:

a new vector that contains the activated elements.

apply

public de.jungblut.math.DoubleMatrix apply(de.jungblut.math.DoubleMatrix matrix)

Description copied from interface: ActivationFunction

Applies the gradient of the activation function on each element in the given matrix.

Specified by:

apply in interface ActivationFunction

Parameters:

matrix - the matrix to apply this function on.

Returns:

a new matrix that contains the gradient of the elements.

gradient

public de.jungblut.math.DoubleVector gradient(de.jungblut.math.DoubleVector vector)

Description copied from interface: ActivationFunction

Applies the gradient of the activation function on each element in the given vector.

Specified by:

gradient in interface ActivationFunction

Parameters:

vector - the vector to apply this function on.

Returns:

a new vector that contains the gradient of the elements.

gradient

public de.jungblut.math.DoubleMatrix gradient(de.jungblut.math.DoubleMatrix matrix)

Description copied from interface: ActivationFunction

Applies the gradient of the activation function on each element in the given matrix.

Specified by:

gradient in interface ActivationFunction

Parameters:

matrix - the matrix to apply this function on.

Returns:

a new matrix that contains the gradient of the elements.

newInstance

protected de.jungblut.math.DoubleMatrix newInstance(de.jungblut.math.DoubleMatrix mat)

newInstance

protected de.jungblut.math.DoubleVector newInstance(de.jungblut.math.DoubleVector v)

toString

public java.lang.String toString()

Overrides:

toString in class java.lang.Object