Class Activation

Represents the activation function that is applied to a neural layer.

Inheritance

Object

Activation

IdentityActivation

ReluActivation

SoftmaxActivation

TanhActivation

Inherited Members

Object.Equals(Object)

Object.Equals(Object, Object)

Object.GetHashCode()

Object.GetType()

Object.MemberwiseClone()

Object.ReferenceEquals(Object, Object)

Object.ToString()

Namespace: NeuralNetLearning.Maths.Activations

Assembly: NeuralNet.dll

Syntax

public abstract class Activation

Methods

Apply(Vector<Double>)

Returns the resulting MathNet.Numerics.LinearAlgebra.Vector<T> on applying the activation function to input.

Declaration

public abstract Vector<double> Apply(Vector<double> input)

Parameters

Type	Name	Description
MathNet.Numerics.LinearAlgebra.Vector<Double>	input

Returns

Type	Description
MathNet.Numerics.LinearAlgebra.Vector<Double>

Derivative(Vector<Double>)

Returns the matrix of derivatives of the activation function evaluated with input.

The [r, c] entry is the derivative of the rth component of the activation function, with respect to the cth component of the input.

Declaration

public abstract Matrix<double> Derivative(Vector<double> input)

Parameters

Type	Name	Description
MathNet.Numerics.LinearAlgebra.Vector<Double>	input

Returns

Type	Description
MathNet.Numerics.LinearAlgebra.Matrix<Double>