#include <neuralNetwork.h>

Inheritance diagram for neuralNetwork< T >:

Collaboration diagram for neuralNetwork< T >:

Public Member Functions
	neuralNetwork (const size_t &num_inputs, const std::vector< size_t > &which_inputs, const size_t &num_hidden_layers, const size_t &num_hidden_nodes, const std::vector< T > &weights, const std::vector< T > &wHiddenOutput, const std::vector< T > &inRanges, const std::vector< T > &inBases, const T &outRange, const T &outBase)

	neuralNetwork (const size_t &num_inputs, const std::vector< size_t > &which_inputs, const size_t &num_hidden_layer, const size_t &num_hidden_nodes)

	~neuralNetwork ()

T	run (const std::vector< T > &inputVector) override

void	reset () override

size_t	getNumInputs () const override

std::vector< size_t >	getWhichInputs () const override

size_t	getNumHiddenLayers () const

void	setNumHiddenLayers (size_t num_hidden_layers)

size_t	getNumHiddenNodes () const

void	setNumHiddenNodes (size_t num_hidden_nodes)

size_t	getEpochs () const

void	setEpochs (const size_t &epochs)

std::vector< T >	getWeights () const

std::vector< T >	getWHiddenOutput () const

std::vector< T >	getInRanges () const

std::vector< T >	getInBases () const

T	getOutRange () const

T	getOutBase () const

void	getJSONDescription (Json::Value &currentModel) override

void	train (const std::vector< trainingExampleTemplate< T > > &trainingSet) override
	These pertain to the training, and aren't need to run a trained model //. More...

void	train (const std::vector< trainingExampleTemplate< T > > &trainingSet, const std::size_t whichOutput) override

size_t	getCurrentEpoch () const

Public Member Functions inherited from baseModel< T >
virtual	~baseModel ()

Additional Inherited Members
Protected Member Functions inherited from baseModel< T >
template<typename TT , class Dummy = int>
Json::Value	vector2json (TT vec)

template<class Dummy = int>
Json::Value	vector2json (std::vector< unsigned long > vec)

Detailed Description

template<typename T>
class neuralNetwork< T >

Class for implementing a Neural Network.

This class includes both running and training, and constructors for reading trained models from JSON.

Constructor & Destructor Documentation

◆ neuralNetwork() [1/2]

template<typename T >

neuralNetwork< T >::neuralNetwork	(	const size_t &	num_inputs,
		const std::vector< size_t > &	which_inputs,
		const size_t &	num_hidden_layers,
		const size_t &	num_hidden_nodes,
		const std::vector< T > &	_weights,
		const std::vector< T > &	w_hidden_output,
		const std::vector< T > &	in_ranges,
		const std::vector< T > &	in_bases,
		const T &	out_range,
		const T &	out_base
	)

This is the constructor for building a trained model from JSON.

This is the constructor for a model imported from JSON.

◆ neuralNetwork() [2/2]

template<typename T >

neuralNetwork< T >::neuralNetwork	(	const size_t &	num_inputs,
		const std::vector< size_t > &	which_inputs,
		const size_t &	num_hidden_layers,
		const size_t &	num_hidden_nodes
	)

This constructor creates a neural network that needs to be trained.

Parameters

num_inputs	is the number of inputs the network will process
which_inputs	is an vector of which values in the input vector are being fed to the network. ex: {0,2,4}
num_hidden_layer	is the number of hidden layers in the network. Must be at least 1.
num_hidden_nodes	is the number of hidden nodes in each hidden layer. Often, this is the same as num_inputs

Returns: A neuralNetwork instance with randomized weights and no normalization values. These will be set or adjusted during training.

This is the constructor for a model that needs to be trained.

◆ ~neuralNetwork()

template<typename T >

neuralNetwork< T >::~neuralNetwork

destructor

This destructor is not needed.

Member Function Documentation

◆ getCurrentEpoch()

template<typename T >

size_t neuralNetwork< T >::getCurrentEpoch

Returns current training epoch.

◆ getEpochs()

template<typename T >

size_t neuralNetwork< T >::getEpochs

◆ getInBases()

template<typename T >

std::vector< T > neuralNetwork< T >::getInBases

◆ getInRanges()

template<typename T >

std::vector< T > neuralNetwork< T >::getInRanges

◆ getJSONDescription()

template<typename T >

void neuralNetwork< T >::getJSONDescription ( Json::Value & currentModel )

overridevirtual

Implements baseModel< T >.

◆ getNumHiddenLayers()

template<typename T >

size_t neuralNetwork< T >::getNumHiddenLayers

◆ getNumHiddenNodes()

template<typename T >

size_t neuralNetwork< T >::getNumHiddenNodes

◆ getNumInputs()

template<typename T >

size_t neuralNetwork< T >::getNumInputs

overridevirtual

Implements baseModel< T >.

◆ getOutBase()

template<typename T >

T neuralNetwork< T >::getOutBase

◆ getOutRange()

template<typename T >

T neuralNetwork< T >::getOutRange

◆ getWeights()

template<typename T >

std::vector< T > neuralNetwork< T >::getWeights

◆ getWhichInputs()

template<typename T >

std::vector< size_t > neuralNetwork< T >::getWhichInputs

overridevirtual

Implements baseModel< T >.

◆ getWHiddenOutput()

template<typename T >

std::vector< T > neuralNetwork< T >::getWHiddenOutput

◆ reset()

template<typename T >

void neuralNetwork< T >::reset

overridevirtual

Implements baseModel< T >.

◆ run()

template<typename T >

T neuralNetwork< T >::run ( const std::vector< T > & inputVector )

overridevirtual

Generate an output value from a single input vector.

Parameters

vector A standard vector of type T that feed-forward regression will run on.

Returns: T A single value, which is the result of the feed-forward operation

Implements baseModel< T >.

◆ setEpochs()

template<typename T >

void neuralNetwork< T >::setEpochs ( const size_t & epochs )

◆ setNumHiddenLayers()

template<typename T >

void neuralNetwork< T >::setNumHiddenLayers ( size_t num_hidden_layers )

◆ setNumHiddenNodes()

template<typename T >

void neuralNetwork< T >::setNumHiddenNodes ( size_t num_hidden_nodes )

◆ train() [1/2]

template<typename T >

void neuralNetwork< T >::train ( const std::vector< trainingExampleTemplate< T > > & trainingSet )

overridevirtual

These pertain to the training, and aren't need to run a trained model //.

Train a model using backpropagation.

Parameters

The	training set is a vector of training examples that contain both a vector of input values and a value specifying desired output.

Implements baseModel< T >.

◆ train() [2/2]

template<typename T >

void neuralNetwork< T >::train	(	const std::vector< trainingExampleTemplate< T > > &	trainingSet,
		const std::size_t	whichOutput
	)

overridevirtual

Train a model using backpropagation. This function is used when the model is part of a modelSet.

Parameters

This	triaining function takes examples that contain a vector of input values and a vector of output values.
The	second argument specifies which output this model is using.

Implements baseModel< T >.

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

src/neuralNetwork.h
src/neuralNetwork.cpp

Public Member Functions

Additional Inherited Members

Detailed Description

template<typename T> class neuralNetwork< T >

Constructor & Destructor Documentation

◆ neuralNetwork() [1/2]

◆ neuralNetwork() [2/2]

◆ ~neuralNetwork()

Member Function Documentation

◆ getCurrentEpoch()

◆ getEpochs()

◆ getInBases()

◆ getInRanges()

◆ getJSONDescription()

◆ getNumHiddenLayers()

◆ getNumHiddenNodes()

◆ getNumInputs()

◆ getOutBase()

◆ getOutRange()

◆ getWeights()

◆ getWhichInputs()

◆ getWHiddenOutput()

◆ reset()

◆ run()

◆ setEpochs()

◆ setNumHiddenLayers()

◆ setNumHiddenNodes()

◆ train() [1/2]

◆ train() [2/2]

template<typename T>
class neuralNetwork< T >