com.antelmann.ai
Class Perceptron

java.lang.Object
  com.antelmann.ai.Perceptron

All Implemented Interfaces:

Serializable

public class Perceptron
extends Object
implements Serializable

Perceptron implements a simple pattern classifier. The Perceptron is an enhanced TLU (Threshold Logic Unit) based on the works by Rosenblatt (Rosenblatt, F. (1962). Principles of Neurodynamics. Spartan Books.). The given implementation here uses boolean values to enforce the duality of the values otherwise represented with 0 and 1. For the purpose of this implementation, false equals 0 and true equals 1. The Perceptron's capability of learning to distinguish two classes (here: not a Java class) of patterns is limited to classes that are linearly separable.

Author:

Holger Antelmann

See Also:

Serialized Form

Constructor Summary

Perceptron(int threshold, int numberOfInputs)

Method Summary

protected Object	clone() overrides Object.clone() to deep-copy the weight array
boolean	compute(boolean[] input)
int	getLength()
int	getThreshold()
int	getWeight(int index)
void	learn(boolean[] input, boolean result) learns a specified output given the input vector by adjusting the synapses (weights) if the result to be learned does not match the actual current output
static void	main(String[] args) testing app
protected void	setThreshold(int threshold)

Methods inherited from class java.lang.Object

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

Constructor Detail

Perceptron

public Perceptron(int threshold,
                  int numberOfInputs)

Method Detail

compute

public boolean compute(boolean[] input)

learn

public void learn(boolean[] input,
                  boolean result)

learns a specified output given the input vector by adjusting the synapses (weights) if the result to be learned does not match the actual current output

setThreshold

protected void setThreshold(int threshold)

getThreshold

public int getThreshold()

getLength

public int getLength()

getWeight

public int getWeight(int index)

clone

protected Object clone()
                throws CloneNotSupportedException

overrides Object.clone() to deep-copy the weight array

Overrides:

clone in class Object

Throws:

CloneNotSupportedException

main

public static void main(String[] args)

testing app