Back to FabImage Library website

You are here: Start » Function Reference » Data Classification » Multilayer Perceptron » MLP_Init

MLP_Init


Header: FIL.h
Namespace: fil
Module: FoundationPro

Creates multilayer perceptron model.

Syntax

C++
C#
 
void fil::MLP_Init
(
	ftl::Optional<const ftl::Array<int>&> inHiddenLayers,
	fil::ActivationFunction::Type inActivationFunction,
	fil::MlpPreprocessing::Type inPreprocessing,
	ftl::Optional<int> inRandomSeed,
	int inInputCount,
	int inOutputCount,
	fil::MlpModel& outMlpModel
)

Parameters

Name Type Range Default Description
Input value
 inHiddenLayers Optional<const Array<int>&> NIL Internal structure of MLP network
Input value
 inActivationFunction ActivationFunction::Type Type of activation function used to calculate neural response
Input value
 inPreprocessing MlpPreprocessing::Type Method of processing input data before learning
Input value
 inRandomSeed Optional<int> 0 - NIL Number used as starting random seed
Input value
 inInputCount int 1 - 1 MLP network input count
Input value
 inOutputCount int 1 - 1 MLP network output count
Output value
 outMlpModel MlpModel& Initialized MlpModel

Description

Filter initializes and sets structure of the MlpModel.

Image: Internal structure of MlpModel. Function f denotes the inActivationFunction.

Parameter inHiddenLayers represents number of neurons in consecutive hidden layers.

The parameter inActivationFunction is a function used to calculate internal neuron activation.

The weights of the multilayer perceptron are initialized by a random numbers. Their values depend on inRandomSeed value.

Parameters inInputCount and inOutputCount defines network inputs and outputs count.

See Also

  • MLP_Train – Creates and trains multilayer perceptron classifier.
  • MLP_Respond – Calculates multilayer perceptron answer.