INeuralNetwork.SetMjuEx

Fore Syntax

SetMjuEx(newMju: Double; layerIndex: Integer);

Fore.NET Syntax

SetMjuEx(newMju: double; layerIndex: integer);

Parameters

newMju. Coefficient of inertia. The value is within the range of [0.0; 1.0].

layerIndex. Layer index.

Description

The SetMjuEx method sets coefficient of inertia of learning for the specified layer of the back-propagation network.

Comments

For details about inertia see INeuralNetwork.SetMju.

Indexation of the network layers is continuous and starts with zero.

Fore Example

As an example, a function is given, which input has back-propagation network fed into (the Net parameter). To execute the example, add a link to the NN system assembly.

Function m_SetMjuNju(Net: NeuralNetwork): NeuralNetwork;
Var
    i: Integer;
Begin
    For i := 0 To Net.GetNumberOfLayers - 1 Do
        Net.SetNjuEx(0.15, i);
        Net.SetMjuEx(0.01, i);
    End For;
Return Net;
End Function m_SetMjuNju;

After executing the example for all the layers of the network there is a unified coefficient of inertia assigned as well as unified learning rate.

Fore.NET Example

As an example, a function is given, which input has back-propagation network fed into (the Net parameter).

Imports Prognoz.Platform.Interop.NN;
…
Public Shared Function m_SetMjuNju(Net: NeuralNetwork): NeuralNetwork;
Var
    i: Integer;
Begin
    For i := 0 To Net.GetNumberOfLayers() - 1 Do
        Net.SetNjuEx(0.15, i);
        Net.SetMjuEx(0.01, i);
    End For;
Return Net;
End Function;

After executing the example a unified coefficient of inertia as well as unified learning rate are assigned for network layers.