C# Класс Encog.Neural.Networks.Training.Lma.JacobianChainRule

Наследование: IComputeJacobian

Показать файл Открыть проект Примеры использования класса

Открытые методы

Метод	Описание
Calculate ( double weights ) : double	Calculate the Jacobian matrix.
JacobianChainRule ( BasicNetwork network, IMLDataSet indexableTraining ) : Encog.Engine.Network.Activation	Construct the chain rule calculation.

Приватные методы

Метод	Описание
CalcDerivative ( IActivationFunction a, double d ) : double	Calculate the derivative.
CalcDerivative2 ( IActivationFunction a, double d ) : double	Calculate the derivative.
CalculateDerivatives ( IMLDataPair pair ) : double	Calculate the derivatives for this training set element.

Описание методов

Calculate() публичный Метод

Calculate the Jacobian matrix.

public Calculate ( double weights ) : double
weights	double	The weights for the neural network.
Результат	double

JacobianChainRule() публичный Метод

Construct the chain rule calculation.

public JacobianChainRule ( BasicNetwork network, IMLDataSet indexableTraining ) : Encog.Engine.Network.Activation
network	Encog.Neural.Networks.BasicNetwork	The network to use.
indexableTraining	IMLDataSet	The training set to use.
Результат	Encog.Engine.Network.Activation