C# Класс numl.Supervised.NeuralNetwork.NetworkTrainingProperties

Network Training Properties class.

Открытые методы

Метод	Описание
Create ( numl.Supervised.NeuralNetwork.Network network, int examples, int features, double learningRate, double lambda, int maxIterations, object parameters = null ) : NetworkTrainingProperties	Creates a new Network Properties object for use in training a neural network model.
NetworkTrainingProperties ( object>.Dictionary parameters = null ) : System	Initializes a new instance of the NetworkTrainingProperties class. Optional custom properties used in training.
this ( string name ) : object	Gets or sets the specified parameter for use in training.

Creates a new Network Properties object for use in training a neural network model.

public static Create ( numl.Supervised.NeuralNetwork.Network network, int examples, int features, double learningRate, double lambda, int maxIterations, object parameters = null ) : NetworkTrainingProperties
network	numl.Supervised.NeuralNetwork.Network	Network being optimized.
examples	int	Number of training examples.
features	int	Number of features in the training examples.
learningRate	double	Learning rate.
lambda	double	Lambda (weight decay).
maxIterations	int	Maximum number of iterations.
parameters	object	Optional parameters object used when training networks. /// Usage: parameters = new { Obj1, Obj2, Obj3... } ///
Результат	NetworkTrainingProperties

Initializes a new instance of the NetworkTrainingProperties class. Optional custom properties used in training.

public NetworkTrainingProperties ( object>.Dictionary parameters = null ) : System
parameters	object>.Dictionary
Результат	System

Gets or sets the specified parameter for use in training.

public this ( string name ) : object
name	string
Результат	object