org.aika.neuron

Class InputNeuron

java.lang.Object

org.aika.neuron.Neuron

org.aika.neuron.InputNeuron

All Implemented Interfaces:

Comparable<Neuron>, Writable

public class InputNeuron
extends Neuron

The InputNeuron class is used to feed information into the neural network. The input neuron has no input synapses. To add or remove input activations simply use the addInput or removeInput methods. An InputNeuron is typically created using the method createOrLookupInputNeuron in the model. By adding an input activation using addInput further activations will be created for this document depending on the network topology. In fact a whole network of neuron activations will be created containing activation objects for all possible interpretations of these inputs. The activations reflect the structure of the activated part of the neural network.

Author:

Lukas Molzberger

Nested Class Summary

Nested classes/interfaces inherited from class org.aika.neuron.Neuron

Neuron.NormWeight

Field Summary

Fields inherited from class org.aika.neuron.Neuron

activationSum, bias, currentNeuronId, id, initialized, inputSynapses, inputSynapsesByWeight, isBlocked, label, LEARN_RATE, lock, m, MAX_SELF_REFERENCING_DEPTH, maxRecurrentSum, negDirSum, negRecSum, node, noTraining, numberOfActivations, outputNodes, outputSynapses, posRecSum, TOLERANCE, WEIGHT_TOLERANCE

Constructor Summary

Constructors

Constructor and Description
InputNeuron()
InputNeuron(String label)
InputNeuron(String label, boolean isBlocked)

Method Summary

Modifier and Type	Method and Description
void	addInput(Document doc, int begin, int end) Propagate an input activation into the network.
void	addInput(Document doc, int begin, int end, Integer rid) Propagate an input activation into the network.
void	addInput(Document doc, int begin, int end, Integer rid, InterprNode o) Propagate an input activation into the network.
void	addInput(Document doc, int begin, int end, Integer rid, InterprNode o, double value) Propagate an input activation into the network.
void	addInput(Document doc, int begin, int end, InterprNode o) Propagate an input activation into the network.
static InputNeuron	create(Document doc, InputNeuron n)
void	propagateAddedActivation(Document doc, Activation act)
void	remove(Document doc)
void	removeInput(Document doc, int begin, int end)
void	removeInput(Document doc, int begin, int end, Integer rid)
void	removeInput(Document doc, int begin, int end, Integer rid, InterprNode o)
void	removeInput(Document doc, int begin, int end, InterprNode o)
String	toString()

Methods inherited from class org.aika.neuron.Neuron

avgActivation, compareTo, computeBounds, computeErrorSignal, computeWeight, count, create, getFinalActivations, propagateRemovedActivation, publish, read, readFields, sigmoid, toStringWithSynapses, train, train, transferFunction, unpublish, write

Methods inherited from class java.lang.Object

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

Constructor Detail

InputNeuron

public InputNeuron()

InputNeuron

public InputNeuron(String label)

InputNeuron

public InputNeuron(String label,
                   boolean isBlocked)

Method Detail

create

public static InputNeuron create(Document doc,
                                 InputNeuron n)

remove

public void remove(Document doc)

Overrides:

remove in class Neuron

propagateAddedActivation

public void propagateAddedActivation(Document doc,
                                     Activation act)

Overrides:

propagateAddedActivation in class Neuron

addInput

public void addInput(Document doc,
                     int begin,
                     int end)

Propagate an input activation into the network.

Parameters:

doc - The current document

begin - The range begin

end - The range end

addInput

public void addInput(Document doc,
                     int begin,
                     int end,
                     InterprNode o)

Propagate an input activation into the network.

Parameters:

doc - The current document

begin - The range begin

end - The range end

o - The interpretation node

addInput

public void addInput(Document doc,
                     int begin,
                     int end,
                     Integer rid)

Propagate an input activation into the network.

Parameters:

doc - The current document

begin - The range begin

end - The range end

rid - The relational id (e.g. the word position)

addInput

public void addInput(Document doc,
                     int begin,
                     int end,
                     Integer rid,
                     InterprNode o)

Propagate an input activation into the network.

Parameters:

doc - The current document

begin - The range begin

end - The range end

rid - The relational id (e.g. the word position)

o - The interpretation node

addInput

public void addInput(Document doc,
                     int begin,
                     int end,
                     Integer rid,
                     InterprNode o,
                     double value)

Propagate an input activation into the network.

Parameters:

doc - The current document

begin - The range begin

end - The range end

rid - The relational id (e.g. the word position)

o - The interpretation node

value - The activation value of this input activation

removeInput

public void removeInput(Document doc,
                        int begin,
                        int end)

removeInput

public void removeInput(Document doc,
                        int begin,
                        int end,
                        InterprNode o)

removeInput

public void removeInput(Document doc,
                        int begin,
                        int end,
                        Integer rid)

removeInput

public void removeInput(Document doc,
                        int begin,
                        int end,
                        Integer rid,
                        InterprNode o)

toString

public String toString()

Overrides:

toString in class Neuron