org.neuroph.nnet.neuron
Class CompetitiveLayer
java.lang.Object
org.neuroph.core.Layer
org.neuroph.nnet.neuron.CompetitiveLayer
- All Implemented Interfaces:
- java.io.Serializable
public class CompetitiveLayer
- extends Layer
The CompetitiveLayer class represents layer of competitive neurons, and it
provides methods for competition.
TODO: competitive learning 3. training dw=n(i-w)
- See Also:
- Serialized Form
Methods inherited from class org.neuroph.core.Layer |
addNeuron, addNeuron, getLayerType, getNeuronAt, getNeuronNum, getNeurons, getNeuronsIterator, indexOf, parentNetwork, randomizeWeights, removeNeuron, removeNeuronAt, reset, setLayerType, setNeuron, setParentNetwork |
Methods inherited from class java.lang.Object |
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait |
serialVersionUID
private static final long serialVersionUID
- See Also:
- Constant Field Values
maxIterations
private int maxIterations
winner
private CompetitiveNeuron winner
CompetitiveLayer
public CompetitiveLayer(int neuronNum,
NeuronProperties neuronProperties)
calculate
public void calculate()
- Performs calculaton for all neurons in this layer
- Overrides:
calculate
in class Layer
getWinner
public CompetitiveNeuron getWinner()
getMaxIterations
public int getMaxIterations()
setMaxIterations
public void setMaxIterations(int maxIterations)