C# Class Accord.Neuro.NguyenWidrow

Nguyen-Widrow weight initialization.

The Nguyen-Widrow initialization algorithm chooses values in order to distribute the active region of each neuron in the layer approximately evenly across the layers' input space.

The values contain a degree of randomness, so they are not the same each time this function is called.

Afficher le fichier Open project: accord-net/framework Class Usage Examples

Méthodes publiques

Méthode	Description
NguyenWidrow ( ActivationNetwork network ) : System	Constructs a new Nguyen-Widrow Weight initialization.
Randomize ( ) : void	Randomizes (initializes) the weights of the network using a Gaussian distribution.
Randomize ( int layerIndex ) : void	Randomizes (initializes) the weights of the network using Nguyen-Widrow method's.

Method Details

NguyenWidrow() public méthode

Constructs a new Nguyen-Widrow Weight initialization.

public NguyenWidrow ( ActivationNetwork network ) : System
network	ActivationNetwork	The activation network whose weights will be initialized.
Résultat	System

Randomize() public méthode

Randomizes (initializes) the weights of the network using a Gaussian distribution.

public Randomize ( ) : void
Résultat	void

Randomize() public méthode

Randomizes (initializes) the weights of the network using Nguyen-Widrow method's.

public Randomize ( int layerIndex ) : void
layerIndex	int
Résultat	void