C# Class SharpNeat.Network.InverseAbsoluteSigmoid

A sigmoid curve produced from the simple and therefore fast arithmetic operations abs, divide and multiply. The curve is designed to match the SteepenedSigmoid activation function as closely as possible.

Inheritance: IActivationFunction

Show file Open project: colgreen/sharpneat

Public Properties

Property	Type	Description
__DefaultInstance	IActivationFunction

Public Methods

Method	Description
Calculate ( double x, double auxArgs ) : double	Calculates the output value for the specified input value and optional activation function auxiliary arguments.
Calculate ( float x, float auxArgs ) : float	Calculates the output value for the specified input value and optional activation function auxiliary arguments. This single precision overload of Calculate() will be used in neural network code that has been specifically written to use floats instead of doubles.
GetRandomAuxArgs ( XorShiftRandom rng, double connectionWeightRange ) : double[]	For activation functions that accept auxiliary arguments; generates random initial values for aux arguments for newly added nodes (from an 'add neuron' mutation).
MutateAuxArgs ( double auxArgs, XorShiftRandom rng, SharpNeat.Utility.ZigguratGaussianSampler gaussianSampler, double connectionWeightRange ) : void	Genetic mutation for auxiliary argument data.

Method Details

Calculate() public method

Calculates the output value for the specified input value and optional activation function auxiliary arguments.

public Calculate ( double x, double auxArgs ) : double
x	double
auxArgs	double
return	double

Calculate() public method

Calculates the output value for the specified input value and optional activation function auxiliary arguments. This single precision overload of Calculate() will be used in neural network code that has been specifically written to use floats instead of doubles.

public Calculate ( float x, float auxArgs ) : float
x	float
auxArgs	float
return	float

GetRandomAuxArgs() public method

For activation functions that accept auxiliary arguments; generates random initial values for aux arguments for newly added nodes (from an 'add neuron' mutation).

public GetRandomAuxArgs ( XorShiftRandom rng, double connectionWeightRange ) : double[]
rng	XorShiftRandom
connectionWeightRange	double
return	double[]

MutateAuxArgs() public method

Genetic mutation for auxiliary argument data.

public MutateAuxArgs ( double auxArgs, XorShiftRandom rng, SharpNeat.Utility.ZigguratGaussianSampler gaussianSampler, double connectionWeightRange ) : void
auxArgs	double
rng	XorShiftRandom
gaussianSampler	SharpNeat.Utility.ZigguratGaussianSampler
connectionWeightRange	double
return	void

Property Details

__DefaultInstance public static property

Default instance provided as a public static field.

public static IActivationFunction __DefaultInstance
return	IActivationFunction