saaps.ann.mlff
Class Neuron

java.lang.Object
  |
  +--saaps.ann.mlff.Neuron

Direct Known Subclasses:

LinNeuron, TanhNeuron

public abstract class Neuron

extends java.lang.Object

This class represents a neuron. The neuron contains the weights and a bias. The class also contains an abstract method for the transfer function. Thus, the class must be inherited by a subclass that implements the transfer function.

Example:

double[] w = {-0.1,0.3,-0.5};
double b = 0.4;
Weights weights = new Weights(w,b);
TanhNeron myNeuron = new TanhNeuron(weights)

See Also:

TanhNeuron, LinNeuron, Layer, Network

Constructor Summary

Neuron()
Creates new empty Neuron.

Neuron(Weights w)
Creates a neuron from a Weights object.

Method Summary

Data	getOutput() Get the output from the neuron.
void	setInput(Data inp) Set the input data to the neuron.
abstract Data	transFun(Data in) The transfer function.

Methods inherited from class java.lang.Object

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

Constructor Detail

Neuron

public Neuron()

Creates new empty Neuron.

Neuron

public Neuron(Weights w)

Creates a neuron from a Weights object.

Parameters:

w - The weights and bias

Method Detail

setInput

public void setInput(Data inp)

Set the input data to the neuron.

Parameters:

inp - The input data

getOutput

public Data getOutput()

Get the output from the neuron.

If new input data has been presented to the neuron then the neuron is first run.

Returns:

The output data

transFun

public abstract Data transFun(Data in)

The transfer function.

Parameters:

in - The data dat goes into the transfer function

Returns:

The output from the tranfer function.