Package me.damoebe.architectures.mlp.structure

Class Neuron

java.lang.Object

me.damoebe.architectures.mlp.structure.Neuron

public class Neuron
extends Object

The class of Neurons used in the Networks

Constructor Summary

Constructors

Constructor	Description
Neuron(List<Connection> connections)	Main constructor of the Neuron class

Method Summary

Modifier and Type	Method	Description
double	getActivation()	Getter for current Activation
List<Connection>	getConnections()	Getter for connections
double	getDelta()	Getter for delta
void	setActivation(double activation)	Sets the current activation (used for neurons in the input layer)
void	setDelta(double delta)	Setter for delta
void	updateActivation()	Updates the current activation based on the connections and sigmoid activation function
void	updateBias(double learningRate)	Updates the bias using learningRate and delta
void	updateWeights(double loss, double noiseRate, double learningRate)	Updates all connection weights based on delta value, loss, noise and learningRate

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Details

Neuron

public Neuron(List<Connection> connections)

Main constructor of the Neuron class

Parameters:

connections - An array featuring all dependencies of the neuron as connections

Method Details

updateActivation

public void updateActivation()

Updates the current activation based on the connections and sigmoid activation function

getActivation

public double getActivation()

Getter for current Activation

Returns:

current activation

setActivation

public void setActivation(double activation)

Sets the current activation (used for neurons in the input layer)

Parameters:

activation - The new activation of the neuron

updateWeights

public void updateWeights(double loss,
 double noiseRate,
 double learningRate)

Updates all connection weights based on delta value, loss, noise and learningRate

Parameters:

loss - The network loss

noiseRate - The noiseRate for the randomizer

learningRate - The learning Rate which is used to update the weights

updateBias

public void updateBias(double learningRate)

Updates the bias using learningRate and delta

Parameters:

learningRate - The learningRate of the network

getConnections

public List<Connection> getConnections()

Getter for connections

Returns:

List of all connections

getDelta

public double getDelta()

Getter for delta

Returns:

the neurons delta

setDelta

public void setDelta(double delta)

Setter for delta

Parameters:

delta - new delta