sann::Network Class Reference

This is the core class, that represents the whole Neural Network. More...

#include <Network.hpp>

Collaboration diagram for sann::Network:

Public Types
typedef std::function< std::vector< double >const std::vector< double > &target, const std::vector< double > &out)>	error_func

Public Member Functions
	Network ()
	Creates a new empty network.
	Network (const std::vector< size_t > &layers, const math::Func &activationFunc, const Layer::weights_initializer &init)
	Network (const std::vector< size_t > &layers, const std::vector< math::Func > &activationFuncs, const Layer::weights_initializer &init)
	Network (const Network &net)
	Creates a new network equal to an existent one. More...
Network &	operator= (const Network &rhs)
	Copy assignment.
Network &	operator= (Network &&rhs)
	Move assignment.
void	setParameters (const sann::parameters &hyperP)
void	setWeights (const std::vector< std::vector< double >> &weights)
void	setWeights (const std::vector< weightsMatrix > &weights)
void	setWeights (std::vector< weightsMatrix > &&weights)
	Sets the weights for each layer of the network using move semantic. More...
void	setRandomWeights ()
void	setErrorFunction (const error_func &error)
	Sets a new error function to minimize. This error function accepts in input the target value and the current output and returns the derivative of the error function. More...
std::vector< weightsMatrix >	getWeights () const
	Returns a vector with the matrix of weights of each layer. More...
std::vector< std::size_t >	getlayersSizes () const
	Returns a vector with the size of each layer. More...
std::vector< double >	compute (const std::vector< double > &inputs)
	Computes the result for the given inputs. More...
void	train (const sann::dataSet &trainingSet, sann::Estimator &est, const sann::parameters &hyperPar)
	Trains the network using the training set passed as input. More...
void	train (const sann::dataSet &trainingSet, const sann::dataSet &testSet, sann::Estimator &trainEst, sann::Estimator &testEst, const sann::parameters &hyperPar)
	Trains the network using the training set passed as input. More...

Static Public Attributes
static sann::Estimator &	nullEstimator = realNullEstimator

Detailed Description

This is the core class, that represents the whole Neural Network.

Constructor & Destructor Documentation

Network()

sann::Network::Network

(

const Network &

net

)

Creates a new network equal to an existent one.

Parameters

net	- The network to copy.

Member Function Documentation

compute()

vector< double > sann::Network::compute

(

const std::vector< double > &

inputs

)

Computes the result for the given inputs.

Parameters

inputs

- The inputs of the network.

Returns

std::vector<double> - The outputs computed by the network.

getlayersSizes()

vector< size_t > sann::Network::getlayersSizes

(

)

const

Returns a vector with the size of each layer.

Returns

vector<size_t> - A vector with the size of each layer.

getWeights()

vector< weightsMatrix > sann::Network::getWeights

(

)

const

Returns a vector with the matrix of weights of each layer.

Returns

vector<weightsMatrix> - The vector with the weights matrix.

setErrorFunction()

void sann::Network::setErrorFunction

(

const error_func &

error

)

Sets a new error function to minimize. This error function accepts in input the target value and the current output and returns the derivative of the error function.

Parameters

error

- The error function to minimize.

setWeights()

void sann::Network::setWeights

(

std::vector< weightsMatrix > &&

weights

)

Sets the weights for each layer of the network using move semantic.

Parameters

weights

- The weights to set.

train() [1/2]

void sann::Network::train	(	const sann::dataSet &	trainingSet,
		sann::Estimator &	est,
		const sann::parameters &	hyperPar
	)

Trains the network using the training set passed as input.

Parameters

trainingSet	- The training set.
est	- The Estimator of the training set.
hyperPar	- The hyperparameters.

train() [2/2]

void sann::Network::train	(	const sann::dataSet &	trainingSet,
		const sann::dataSet &	testSet,
		sann::Estimator &	trainEst,
		sann::Estimator &	testEst,
		const sann::parameters &	hyperPar
	)

Trains the network using the training set passed as input.

Parameters

trainingSet	- The training set.
testSet	- The test set.
trainEst	- The Estimator for the training set.
testEst	- The Estimator for the test set.
hyperPar	- The hyperparameters.

---

The documentation for this class was generated from the following files:

• sann/Network.hpp
• sann/Network.cpp