C# Класс numl.Supervised.NeuralNetwork.Recurrent.RecurrentNeuron

An individual Gated Recurrent Neuron

Наследование: Neuron

Открытые методы

Метод	Описание
Error ( double t, NetworkTrainingProperties properties ) : double	Returns the error given the supplied error derivative.
Evaluate ( ) : double	Evaluates the state.
RecurrentNeuron ( ) : System	Initializes a new Recurrent Neuron.
Reset ( NetworkTrainingProperties properties ) : void	Resets the state of the current neuron.
Update ( NetworkTrainingProperties properties ) : void	Updates the weights using the supplied (NetworkTrainingProperties)

Returns the error given the supplied error derivative.

public Error ( double t, NetworkTrainingProperties properties ) : double
t	double	The error from the next layer.
properties	NetworkTrainingProperties	Network training properties object.
Результат	double

Evaluates the state.

public Evaluate ( ) : double
Результат	double

Initializes a new Recurrent Neuron.

public RecurrentNeuron ( ) : System
Результат	System

Resets the state of the current neuron.

public Reset ( NetworkTrainingProperties properties ) : void
properties	NetworkTrainingProperties	Network training properties.
Результат	void

Updates the weights using the supplied (NetworkTrainingProperties)

public Update ( NetworkTrainingProperties properties ) : void
properties	NetworkTrainingProperties	Network training properties.
Результат	void