Neuron Class Reference

#include <Neuron.h>

Inheritance diagram for Neuron:

Public Member Functions
	Neuron (std::function< double(double)> activationFunc, std::vector< Edge * > inputs, std::vector< Edge * > outputs)
	Neuron (std::function< double(double)> activationFunc)
void	setInputs (std::vector< Edge * > inputs)
void	setOutputs (std::vector< Edge * > outputs)
void	calcOutput ()
void	calcDelta ()
double	getOutput ()
double	getDelta ()

Protected Member Functions
double	sigmoidDelta ()

Protected Attributes
double	delta = 0

Detailed Description

A neuron, part of a neural network.

Constructor & Destructor Documentation

Neuron::Neuron	(	std::function< double(double)>	activationFunc,
		std::vector< Edge * >	inputs,
		std::vector< Edge * >	outputs
	)

Create a new Neuron with the given in- and outputs

Parameters

activationFunc	function to use for neuron activation
inputs	vector of pointers to incoming edges
outputs	vector of pointers to outgoing edges

Neuron::Neuron

(

std::function< double(double)>

activationFunc

)

Create a new Neuron

Parameters

activationFunc

function to use for neuron activation

Member Function Documentation

void Neuron::calcDelta

(

)

Calculate and store the delta from this Neuron according to the GDR. All output Neurons should have their deltas updated before calculating this.

void Neuron::calcOutput

(

)

Calculate and store the output from this Neuron. All input Neurons should have their outputs updated before calculating this.

double Neuron::getDelta

(

)

Get delta for backpropagation

Returns

the delta of the Neuron

double Neuron::getOutput ( )

virtual

Get the output of this Neuron
Inherits:
Get the output of the node

Returns

the node's output

Implements Node.

void Neuron::setInputs

(

std::vector< Edge * >

inputs

)

Set the edges in to this Neuron

Parameters

inputs

vector of pointers to incoming edges

void Neuron::setOutputs

(

std::vector< Edge * >

outputs

)

Set the edges out from this Neuron

Parameters

outputs

vector of pointers to outgoing edges

double Neuron::sigmoidDelta ( )

protected

Get the derivative of a sigmoid activation function for this node

Returns

the derivative of the sigmoid

Member Data Documentation

double Neuron::delta = 0

protected

Delta for backpropagation

---

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

• src/netbuilding/Neuron.h
• src/netbuilding/Neuron.cpp