All Packages  Class Hierarchy  This Package  Previous  Next  Index

Class BrainWave.SOM.SOMLearn

java.lang.Object
   |
   +----java.awt.Component
           |
           +----BrainWave.BWObject
                   |
                   +----BrainWave.BWButton
                           |
                           +----BrainWave.SOM.SOMLearn
public class SOMLearn
extends BWButton
Implements the SOM learning algorithm. A SOM network has a set of input units and a set of map units. For each input pattern:
  1. Input is instantiated on the input units.
  2. The map unit with the weight vector that is closest to the input pattern becomes the winner.
  3. The activations of the map units are updated according to the weight between the winner unit and that unit.
  4. The weight vectors of the map units are moved towards the input pattern. The amount of movement depends on how active that map unit has become.
Version:
2.0
Author:
Simon Dennis

Constructor Index

o SOMLearn()

Method Index

o actionPerformed(ActionEvent)
o activate()
o getNumberOfCycles()
o getTotalCycles()
o setNumberOfCycles(int)
o setTotalCycles(int)
o setupMenu()

Constructors

o SOMLearn 
 public SOMLearn()

Methods

o getNumberOfCycles 
 public int getNumberOfCycles()
o setNumberOfCycles 
 public void setNumberOfCycles(int i)
o getTotalCycles 
 public int getTotalCycles()
o setTotalCycles 
 public void setTotalCycles(int i)
o activate 
 public void activate()
Overrides:
activate in class BWButton
o setupMenu 
 public void setupMenu()
Overrides:
setupMenu in class BWButton
o actionPerformed 
 public void actionPerformed(ActionEvent evt)
Overrides:
actionPerformed in class BWButton
All Packages  Class Hierarchy  This Package  Previous  Next  Index

On This Site