INeuron (aika 0.9 API)

java.lang.Object
- org.aika.AbstractNode<Neuron>
- - org.aika.neuron.INeuron

All Implemented Interfaces:

Comparable<INeuron>, Writable
```
public class INeuron
extends AbstractNode<Neuron>
implements Comparable<INeuron>
```
The INeuron class represents a internal neuron implementation in Aikas neural network and is connected to other neurons through input synapses and output synapses. The activation value of a neuron is calculated by computing the weighted sum (input act. value * synapse weight) of the input synapses, adding the bias to it and sending the resulting value through a transfer function (the upper part of tanh).

The neuron does not store its activations by itself. The activation objects are stored within the logic nodes. To access the activations of this neuron simply use the member variable node or use the method getFinalActivations(Document doc) to ge the final activations of this neuron.

Author:

Lukas Molzberger

Nested Class Summary

Nested Classes
Modifier and Type Class and Description

static class INeuron.NormWeight

Nested Classes
Modifier and Type	Class and Description
`static class`	`INeuron.NormWeight`

Field Summary

Fields
Modifier and Type	Field and Description
`double`	`activationSum`
`double`	`bias`
`TreeMap<Synapse,Synapse>`	`inputSynapses`
`boolean`	`isBlocked`
`String`	`label`
`ReadWriteLock`	`lock`
`static int`	`MAX_SELF_REFERENCING_DEPTH`
`double`	`maxRecurrentSum`
`double`	`negDirSum`
`double`	`negRecSum`
`Provider<OrNode>`	`node`
`boolean`	`noTraining`
`int`	`numberOfActivations`
`TreeMap<Synapse.Key,Provider<InputNode>>`	`outputNodes`
`double`	`posRecSum`
`static double`	`TOLERANCE`
`static double`	`WEIGHT_TOLERANCE`

Fields inherited from class org.aika.AbstractNode
lastUsedDocumentId, provider

Constructor Summary

Constructors
Constructor and Description
`INeuron(Model m)`
`INeuron(Model m, String label)`
`INeuron(Model m, String label, boolean isBlocked, boolean noTraining)`

Method Summary

All Methods Static Methods Instance Methods Concrete Methods
Modifier and Type	Method and Description
`Activation`	`addInput(Document doc, int begin, int end, Integer rid, InterprNode o, double value)` Propagate an input activation into the network.
`static INeuron`	`addSynapse(Model m, int threadId, Neuron pn, double biasDelta, Synapse s)`
`double`	`avgActivation()`
`int`	`compareTo(INeuron n)`
`void`	`computeBounds(Activation act)`
`void`	`computeErrorSignal(Document doc, Activation act)`
`Activation.State`	`computeWeight(int round, Activation act, SearchNode sn, Document doc)`
`void`	`count(Document doc)`
`Collection<Activation>`	`getFinalActivations(Document doc)` `getFinalActivations` is a convenience method to retrieve all activations of the given neuron that are part of the final interpretation.
`Synapse`	`getInputSynapse(Synapse s)`
`static Neuron`	`init(Model m, int threadId, Neuron pn, double bias, Set<Synapse> inputs)`
`void`	`linkNeuronRelations(Document doc, Activation act)` Sets the incoming and outgoing links between neuron activations.
`void`	`propagateAddedActivation(Document doc, Activation act)`
`void`	`propagateRemovedActivation(Document doc, NodeActivation act)`
`void`	`publish()`
`void`	`reactivate()`
`void`	`readFields(DataInput in, Model m)` Deserialize the fields of this object from `in`.
`static INeuron`	`readNeuron(DataInput in, Neuron p)`
`void`	`remove()`
`void`	`removeInput(Document doc, int begin, int end, Integer rid, InterprNode o)`
`static double`	`sigmoid(double x)`
`void`	`suspend()`
`String`	`toString()`
`String`	`toStringWithSynapses()`
`void`	`train(Document doc, Activation targetAct, double learnRate, TrainConfig.SynapseEvaluation se)`
`static double`	`transferFunction(double x)`
`static void`	`unlinkNeuronRelations(Document doc, Activation act)`
`void`	`unpublish()`
`void`	`write(DataOutput out)` Serialize the fields of this object to `out`.

Methods inherited from class org.aika.AbstractNode
read

Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Field Detail

WEIGHT_TOLERANCE
```
public static double WEIGHT_TOLERANCE
```

TOLERANCE
```
public static double TOLERANCE
```

MAX_SELF_REFERENCING_DEPTH

public static int MAX_SELF_REFERENCING_DEPTH

label
```
public String label
```

bias
```
public volatile double bias
```

negDirSum
```
public volatile double negDirSum
```

negRecSum
```
public volatile double negRecSum
```

posRecSum
```
public volatile double posRecSum
```

maxRecurrentSum
```
public volatile double maxRecurrentSum
```

inputSynapses

public TreeMap<Synapse,Synapse> inputSynapses

outputNodes

public TreeMap<Synapse.Key,Provider<InputNode>> outputNodes

node
```
public Provider<OrNode> node
```

isBlocked
```
public boolean isBlocked
```

noTraining
```
public boolean noTraining
```

activationSum
```
public volatile double activationSum
```

numberOfActivations

public volatile int numberOfActivations

lock
```
public ReadWriteLock lock
```

Constructor Detail

INeuron
```
public INeuron(Model m)
```

INeuron

public INeuron(Model m,
               String label)

INeuron

public INeuron(Model m,
               String label,
               boolean isBlocked,
               boolean noTraining)

Method Detail

addInput

public Activation 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,
                        Integer rid,
                        InterprNode o)

avgActivation
```
public double avgActivation()
```

publish
```
public void publish()
```

unpublish
```
public void unpublish()
```

remove
```
public void remove()
```

propagateAddedActivation

public void propagateAddedActivation(Document doc,
                                     Activation act)

propagateRemovedActivation

public void propagateRemovedActivation(Document doc,
                                       NodeActivation act)

computeBounds

public void computeBounds(Activation act)

computeWeight

public Activation.State computeWeight(int round,
                                      Activation act,
                                      SearchNode sn,
                                      Document doc)

computeErrorSignal

public void computeErrorSignal(Document doc,
                               Activation act)

train

public void train(Document doc,
                  Activation targetAct,
                  double learnRate,
                  TrainConfig.SynapseEvaluation se)

transferFunction

public static double transferFunction(double x)

sigmoid

public static double sigmoid(double x)

count
```
public void count(Document doc)
```

linkNeuronRelations

public void linkNeuronRelations(Document doc,
                                Activation act)

Sets the incoming and outgoing links between neuron activations.

Parameters:: doc -; act -

unlinkNeuronRelations

public static void unlinkNeuronRelations(Document doc,
                                         Activation act)

getInputSynapse

public Synapse getInputSynapse(Synapse s)

compareTo
```
public int compareTo(INeuron n)
```
Specified by:

compareTo in interface Comparable<INeuron>

write
```
public void write(DataOutput out)
           throws IOException
```
Description copied from interface: Writable

Serialize the fields of this object to out.

Specified by:

write in interface Writable

Parameters:

out - DataOuput to serialize this object into.

Throws:

IOException

readFields
```
public void readFields(DataInput in,
                       Model m)
                throws IOException
```
Description copied from interface: Writable

Deserialize the fields of this object from in.
For efficiency, implementations should attempt to re-use storage in the existing object where possible.

Specified by:

readFields in interface Writable

Parameters:

in - DataInput to deseriablize this object from.

Throws:

IOException

suspend
```
public void suspend()
```
Overrides:

suspend in class AbstractNode<Neuron>

reactivate
```
public void reactivate()
```
Overrides:

reactivate in class AbstractNode<Neuron>

init

public static Neuron init(Model m,
                          int threadId,
                          Neuron pn,
                          double bias,
                          Set<Synapse> inputs)

addSynapse

public static INeuron addSynapse(Model m,
                                 int threadId,
                                 Neuron pn,
                                 double biasDelta,
                                 Synapse s)

readNeuron

public static INeuron readNeuron(DataInput in,
                                 Neuron p)
                          throws IOException

Throws:: IOException

toString
```
public String toString()
```
Overrides:

toString in class Object

toStringWithSynapses
```
public String toStringWithSynapses()
```

getFinalActivations
```
public Collection<Activation> getFinalActivations(Document doc)
```
getFinalActivations is a convenience method to retrieve all activations of the given neuron that are part of the final interpretation. Before calling this method, the doc.process() needs to be called first.

Parameters:

doc - The current document

Returns:

A collection with all final activations of this neuron.

Class INeuron

Nested Class Summary

Field Summary

Fields inherited from class org.aika.AbstractNode

Constructor Summary

Method Summary

Methods inherited from class org.aika.AbstractNode

Methods inherited from class java.lang.Object

Field Detail

WEIGHT_TOLERANCE

TOLERANCE

MAX_SELF_REFERENCING_DEPTH

label

bias

negDirSum

negRecSum

posRecSum

maxRecurrentSum

inputSynapses

outputNodes

node

isBlocked

noTraining

activationSum

numberOfActivations

lock

Constructor Detail

INeuron

INeuron

INeuron

Method Detail

addInput

removeInput

avgActivation

publish

unpublish

remove

propagateAddedActivation

propagateRemovedActivation

computeBounds

computeWeight

computeErrorSignal

train

transferFunction

sigmoid

count

linkNeuronRelations

unlinkNeuronRelations

getInputSynapse

compareTo

write

readFields

suspend

reactivate

init

addSynapse

readNeuron

toString

toStringWithSynapses

getFinalActivations