Overview  Package  Class  Use  Tree  Deprecated  Index  Help 
 PREV   NEXT FRAMES    NO FRAMES    

Serialized Form

Package org.joone.engine

Class org.joone.engine.AbstractLearner extends java.lang.Object implements Serializable

Serialized Fields

learnable

Learnable learnable
The learnable, the object that is subjected to the learning process.

learnableLayer

LearnableLayer learnableLayer
The layer (biases) that is subjected to the learning process.

learnableSynapse

LearnableSynapse learnableSynapse
The synapse (weights) that is subjected to the learning process.

monitor

Monitor monitor
The saved monitor object.

Class org.joone.engine.BasicLearner extends ExtendableLearner implements Serializable

Class org.joone.engine.BatchLearner extends ExtendableLearner implements Serializable

Class org.joone.engine.BiasedLinearLayer extends SimpleLayer implements Serializable

Class org.joone.engine.BufferedSynapse extends Synapse implements Serializable

serialVersionUID: -8067243400677466498L

Class org.joone.engine.CircularSpatialMap extends SpatialMap implements Serializable

serialVersionUID: 442118480555350769L

Class org.joone.engine.ContextLayer extends SimpleLayer implements Serializable

serialVersionUID: -8773800970295287404L

Serialized Fields

beta

double beta

timeConstant

double timeConstant

initialState

Matrix initialState
The initial state

Class org.joone.engine.DelayLayer extends MemoryLayer implements Serializable

serialVersionUID: 1547734529107850525L

Class org.joone.engine.DelaySynapse extends Synapse implements Serializable

serialVersionUID: 8268129000639124340L

Serialized Fields

fir

FIRFilter[][] fir

taps

int taps

Class org.joone.engine.DirectSynapse extends Synapse implements Serializable

serialVersionUID: 3079898042708755094L

Class org.joone.engine.EKFFFNLearnerPlugin extends MonitorPlugin implements Serializable

serialVersionUID: 1L

Serialized Fields

kalmanBoost

double kalmanBoost
The kalman boost factor

verbose

int verbose
Zero for no verbosity, otherwise the number of patterns after which we croak

processorCount

int processorCount
Number of processors to use, 2 or more uses SMP

Class org.joone.engine.EKFRNNLearnerPlugin extends MonitorPlugin implements Serializable

serialVersionUID: 1L

Serialized Fields

kalmanBoost

double kalmanBoost
The kalman boost factor

verbose

int verbose
Zero for no verbosity, otherwise the number of patterns after which we croak

processorCount

int processorCount
Number of processors to use, if 2 or more, will use SMP

Class org.joone.engine.ExtendableLearner extends AbstractLearner implements Serializable

Serialized Fields

theDeltaRuleExtenders

java.util.List<E> theDeltaRuleExtenders
The list with delta rule extenders, extenders that change the delta w, e.g. momentum term, etc.

theGradientExtenders

java.util.List<E> theGradientExtenders
The list with gradient extenders, extenders that change the gradient.

theUpdateWeightExtender

UpdateWeightExtender theUpdateWeightExtender
The update weight extender, that is, the way to update the weights, online, batch mode, etc.

Class org.joone.engine.Fifo extends java.util.Vector implements Serializable

serialVersionUID: -3937649024771901836L

Class org.joone.engine.FIRFilter extends java.lang.Object implements Serializable

serialVersionUID: 2539307324689626619L

Serialized Fields

m_taps

int m_taps

memory

double[] memory

backmemory

double[] backmemory

outs

double[] outs

bouts

double[] bouts

array

Matrix array

lrate

double lrate

momentum

double momentum

Class org.joone.engine.FreudRuleFullSynapse extends FullSynapse implements Serializable

serialVersionUID: 4391516546875376355L

Class org.joone.engine.FullSynapse extends Synapse implements Serializable

serialVersionUID: 5518898101307425554L

Class org.joone.engine.GaussianLayer extends SimpleLayer implements Serializable

serialVersionUID: -941653911909171430L

Serialized Fields

LayerWidth

int LayerWidth

LayerHeight

int LayerHeight

LayerDepth

int LayerDepth

space_map

SpatialMap space_map

timeConstant

double timeConstant

orderingPhase

int orderingPhase

initialGaussianSize

double initialGaussianSize

Class org.joone.engine.GaussianSpatialMap extends SpatialMap implements Serializable

serialVersionUID: -5578079370364572387L

Class org.joone.engine.GaussLayer extends SimpleLayer implements Serializable

Class org.joone.engine.KohonenSynapse extends FullSynapse implements Serializable

serialVersionUID: -4966435217407942471L

Serialization Methods

readObject

private void readObject(java.io.ObjectInputStream in)
                 throws java.io.IOException,
                        java.lang.ClassNotFoundException
Throws:
java.io.IOException
java.lang.ClassNotFoundException
Serialized Fields

currentLearningRate

double currentLearningRate

timeConstant

double timeConstant

orderingPhase

int orderingPhase

Class org.joone.engine.Layer extends java.lang.Object implements Serializable

serialVersionUID: -1572591602454639355L

Serialization Methods

readObject

private void readObject(java.io.ObjectInputStream in)
                 throws java.io.IOException,
                        java.lang.ClassNotFoundException
Read in a serialised version of this layer

Throws:
java.io.IOException
java.lang.ClassNotFoundException

writeObject

private void writeObject(java.io.ObjectOutputStream out)
                  throws java.io.IOException
Write a serialized version of this layer

Throws:
java.io.IOException
Serialized Fields

LayerName

java.lang.String LayerName
The name of the layer

rows

int rows
The number of neurons in the layer

bias

Matrix bias
Holds the bias of neurons of the layer

monitor

Monitor monitor
The monitor of the layer. Contains all parameters needed to the learning phase

m_batch

int m_batch
Not used but maintained for backward serialization compatability.

learning

boolean learning
The Net's phase: false == recall; true == learning

learnable

boolean learnable
Contains true if for the Layer must be used a Learner instead of a built-in learning algorithm. Set it in the constructor of any inherited class. Used by the getLearner method.

See Also:
getLearner

inputPatternListeners

java.util.Vector<E> inputPatternListeners
Contains the list of input connected listeners (InputPatternListener)

outputPatternListeners

java.util.Vector<E> outputPatternListeners
Contains the list of output connected listeners (OutputPatternListener)

Class org.joone.engine.LinearLayer extends SimpleLayer implements Serializable

serialVersionUID: 2243109263560495304L

Serialized Fields

beta

double beta

Class org.joone.engine.LogarithmicLayer extends SimpleLayer implements Serializable

serialVersionUID: -4983197905588348060L

Class org.joone.engine.Matrix extends java.lang.Object implements Serializable

serialVersionUID: -1392966842649908366L

Serialization Methods

readObject

private void readObject(java.io.ObjectInputStream in)
                 throws java.io.IOException,
                        java.lang.ClassNotFoundException
Throws:
java.io.IOException
java.lang.ClassNotFoundException
Serialized Fields

value

double[][] value
The values of the weights / biases.

delta

double[][] delta
The value of the last modification, i.e. the last update.

enabled

boolean[][] enabled
Flag indicating whether the weight is on or off.

fixed

boolean[][] fixed
Flag indicating whether the weight is fixed or trainable / adjustable.

m_rows

int m_rows
The number of rows. That is, in case of weights, the number of neurons on the input side of the synapse. In case of biases, the number of neurons.

m_cols

int m_cols
The number of columns. That is, in case of weights, the number of neurons on the output side of the synapse. In case of biases, the value equals 0.

weightInitializer

WeightInitializer weightInitializer
The weight initializer that is used by this class.

Class org.joone.engine.MemoryLayer extends Layer implements Serializable

serialVersionUID: 5447777678414684948L

Serialization Methods

readObject

private void readObject(java.io.ObjectInputStream in)
                 throws java.io.IOException,
                        java.lang.ClassNotFoundException
Throws:
java.io.IOException
java.lang.ClassNotFoundException
Serialized Fields

memory

double[] memory

backmemory

double[] backmemory

taps

int taps

Class org.joone.engine.Monitor extends java.lang.Object implements Serializable

serialVersionUID: 2909501813894146845L

Serialization Methods

readResolve

protected java.lang.Object readResolve()
Serialized Fields

preLearning

int preLearning

learning

boolean learning

currentCicle

int currentCicle

run

int run

saveCurrentCicle

int saveCurrentCicle

saveRun

int saveRun

patterns

int patterns

validationPatterns

int validationPatterns

totCicles

int totCicles

learningRate

double learningRate

momentum

double momentum

globalError

double globalError

batchSize

int batchSize

useRMSE

boolean useRMSE
Use RMSE (if true) for back propagation, MSE (if false) otherwise.

theLearnerFactory

LearnerFactory theLearnerFactory
The learner factory. If set this factory provides synapses and layers with learners.

parent

Monitor parent

supervisioned

boolean supervisioned

singleThreadMode

boolean singleThreadMode

learningMode

int learningMode

learners

java.util.List<E> learners

params

java.util.Hashtable<K,V> params

backprop

boolean backprop
Ability to switch on and off the error backpropagation method, specifically for use with RTRL

Class org.joone.engine.NeuralNetEvent extends java.util.EventObject implements Serializable

serialVersionUID: -2307998901508765401L

Serialized Fields

nnet

NeuralNet nnet

Class org.joone.engine.OutputSwitchSynapse extends java.lang.Object implements Serializable

serialVersionUID: 2906294213180089226L

Serialized Fields

outputs

java.util.Vector<E> outputs

name

java.lang.String name

mon

Monitor mon

inputDimension

int inputDimension

outputFull

boolean outputFull

enabled

boolean enabled

activeSynapse

OutputPatternListener activeSynapse

defaultSynapse

OutputPatternListener defaultSynapse

Class org.joone.engine.Pattern extends java.lang.Object implements Serializable

serialVersionUID: -609786590797838450L

Serialized Fields

count

int count

array

double[] array

outArray

double[] outArray

Class org.joone.engine.RbfGaussianLayer extends RbfLayer implements Serializable

Serialized Fields

theGaussianParameters

RbfGaussianParameters[] theGaussianParameters
The parameters for the different Gaussian RBFs (neurons)

theUseRandomSelector

boolean theUseRandomSelector
Flag indication if we should use randomly chosen fixed centers.

theRandomSelector

RbfRandomCenterSelector theRandomSelector
The random selector (if theUseRandomSelector equals true).

Class org.joone.engine.RbfGaussianParameters extends java.lang.Object implements Serializable

Serialized Fields

theMean

double[] theMean
The mean (center) of the RBF.

theStdDeviation

double theStdDeviation
The standard deviation (sigma).

Class org.joone.engine.RbfInputSynapse extends Synapse implements Serializable

Class org.joone.engine.RbfLayer extends Layer implements Serializable

Class org.joone.engine.RpropLearner extends ExtendableLearner implements Serializable

Serialized Fields

theRpropExtender

RpropExtender theRpropExtender
The RPROP extender. Only used to make back compatibility possible.

Class org.joone.engine.RTRLLearnerFactory extends java.lang.Object implements Serializable

Serialized Fields

lineseek

boolean lineseek
True if the experimental lineseek version is in use

learners

java.util.List<E> learners
List of all learners referenced here, not used per se

z

java.util.List<E> z
The nodes, referred to as the z element

Z

java.util.List<E> Z
The same as z, but this time ordered so that nodes in U come first. We are more interested in Z, but use z to extract the nodes and then at a later stage sort it into Z.

T

java.util.List<E> T
The output nodes, a subset of Z and used to speed up calculations

U

java.util.List<E> U
The nodes in U, a subset of Z, specifically the first few nodes in Z, but those that are not input nodes.

inputLayer

Layer inputLayer
The input layer

outputLayer

Layer outputLayer
The output layer from which we calculate errors and which we use to determine if a node is in T

learningRate

double learningRate
The learning rate, stored locally to speed things up and adjusted dynamically if the experimental lineseek version is used.

momentum

double momentum
The momentum, stored locally to speed things up

p

double[][][] p
The p matrix. It is a 3d matrix with dimensions the number of nodes in U x number of nodes in U x number of nodes in Z, typically indexed via K. In the literature this is typically indexed via sup k, sub i and j.

updateP

double[][][] updateP
A utility matrix used to speed up things and when we update p. We will swap this with p when we update.

q

double[][] q
The q matrix. This is a 2d matrix, but very similar to the p matrix. It is used for initial states and not weights though. For now, we only optimise the weights and not the initial state, so it is not used at present.

updateQ

double[][] updateQ
A utility matrix used to speed up things and when we update q. We will swap this with q when we update. As with q, it is not used at present.

lastError

double[] lastError
The error last seen at the output layer T

network

NeuralNet network
The neural network trained here

currentSSE

double currentSSE
The current cycle's sum of squared errors, used especially when the experimental lineseek version is enabled

previousSSE

double previousSSE
The previous cycle's sum of squared errors, used especially when the experimental lineseek version is enabled

stepUpScale

double stepUpScale
The scaling to apply if we are stepping along the gradient, used when the experimental lineseek version is enabled

stepDownScale

double stepDownScale
The scaling to apply if we are stepping down along an updated gradient, used when the experimental lineseek version is enabled

verbose

boolean verbose
If true, display progress along the way

upperLearningRate

double upperLearningRate
Upper bound for learning rate

lowerLearningRate

double lowerLearningRate
Lower bound for learning rate

patternCount

int patternCount
The current pattern. Not really required, sometimes used for debugging purposes.

random

java.util.Random random
A random number generator used to shock the weights

shockFactor

double shockFactor
The shock factor

weightMagnitude

double weightMagnitude
The maximum weight magnitude. Weights are not allowed to exceed this magnitude.

interCycleUpdates

boolean interCycleUpdates
True if we regularly update the weights within a cycle, false if we only update once the cycle completes

minimumPatternCount

int minimumPatternCount
Number of patterns that have to be seen before weights are updated

updateProbability

double updateProbability
Probability of updating weights at next pattern

weights

java.util.List<E> weights
List of all weights, used when setting up a CFOProblem

cycleCount

int cycleCount
Evaluation count, used when setting up a CFOProblem

Class org.joone.engine.RTRLLearnerFactory.InitialState extends java.lang.Object implements Serializable

Serialized Fields

layer

ContextLayer layer
The context layer that this state is associated with

index

int index
The index into the layer at which this initial state is found

k

int k
The node in z into which this state will be fired

Class org.joone.engine.RTRLLearnerFactory.Node extends java.lang.Object implements Serializable

Serialized Fields

bias

double bias
The bias if this is a bias node

layer

Layer layer
The layer at which this node is found

index

int index
The index into the layer at which this node is found

K

int K
The node's number in Z and U if it is in U

inU

boolean inU
True if this node is in U, false if not

weights

java.util.List<E> weights
The list of weights that carry a signal into this node

I

java.util.List<E> I
The list of weights that carry a signal into this node from a node in U

initialStates

java.util.List<E> initialStates
The list of initial states that fire into this node, not used at present but maybe for a future version in which the state is also optimised.

Class org.joone.engine.RTRLLearnerFactory.RTRLLearner extends java.lang.Object implements Serializable

Serialized Fields

learnable

Learnable learnable
The learnable that is trained here

Class org.joone.engine.RTRLLearnerFactory.Weight extends java.lang.Object implements Serializable

Serialized Fields

K

int K
The node K in Z from which this weight comes

i

int i
The joone i index of this weight

j

int j
The joone j index of this weight

layer

Layer layer
The joone layer which is used if this weight is a bias

synapse

Synapse synapse
The joone synapse which is used if this weight is a normal synaptical weight

Class org.joone.engine.RTRLLearnerPlugin extends MonitorPlugin implements Serializable

serialVersionUID: 1L

Serialized Fields

processorCount

int processorCount
Number of processors to use, 1 or less on uniprocessors

Class org.joone.engine.SangerSynapse extends FullSynapse implements Serializable

serialVersionUID: 1417085683178232377L

Class org.joone.engine.SigmoidLayer extends SimpleLayer implements Serializable

serialVersionUID: -8700747963164046048L

Serialized Fields

flatSpotConstant

double flatSpotConstant
Constant to overcome the "flat spot" problem. This problem is described in: S.E. Fahlman, "An emperical study of learning speed in backpropagation with good scaling properties," Dept. Comput. Sci. Carnegie Mellon Univ., Pittsburgh, PA, Tech. Rep., CMU-CS-88-162, 1988. Setting this constant to 0 (default value), the derivative of the sigmoid function is unchanged (normal function). An good value for this constant might be 0.1.

Class org.joone.engine.SimpleLayer extends Layer implements Serializable

serialVersionUID: -2579073586181182767L

Serialized Fields

lrate

double lrate

momentum

double momentum

Class org.joone.engine.SineLayer extends SimpleLayer implements Serializable

serialVersionUID: -2636086679111635756L

Class org.joone.engine.SoftmaxLayer extends LinearLayer implements Serializable

serialVersionUID: 2243109263560495355L

Class org.joone.engine.SpatialMap extends java.lang.Object implements Serializable

Serialized Fields

InitialGaussianSize

double InitialGaussianSize

CurrentGaussianSize

double CurrentGaussianSize

map_width

int map_width

map_height

int map_height

map_depth

int map_depth

win_x

int win_x

win_y

int win_y

win_z

int win_z

TotalEpochs

int TotalEpochs

orderingPhase

int orderingPhase

TimeConstant

double TimeConstant

Class org.joone.engine.Synapse extends java.lang.Object implements Serializable

serialVersionUID: -5892822057908231022L

Serialization Methods

readResolve

protected java.lang.Object readResolve()
Serialized Fields

fieldName

java.lang.String fieldName
Name set by default to "Synapse [synapse count]"

learningRate

double learningRate

momentum

double momentum

inputDimension

int inputDimension

outputDimension

int outputDimension

inputFull

boolean inputFull

outputFull

boolean outputFull

monitor

Monitor monitor

ignoreBefore

int ignoreBefore

loopBack

boolean loopBack

array

Matrix array

m_batch

int m_batch

enabled

boolean enabled

learnable

boolean learnable
Contains true if for the current Synapse must be used a Learner instead of a built-in learning algorithm. Set it in the constructor of any inherited class. Used by the getLearner method.

See Also:
getLearner

Class org.joone.engine.TanhLayer extends SimpleLayer implements Serializable

serialVersionUID: -2073914754873517298L

Serialized Fields

flatSpotConstant

double flatSpotConstant
Constant to overcome the "flat spot" problem. This problem is described in: S.E. Fahlman, "An emperical study of learning speed in backpropagation with good scaling properties," Dept. Comput. Sci. Carnegie Mellon Univ., Pittsburgh, PA, Tech. Rep., CMU-CS-88-162, 1988. Setting this constant to 0 (default value), the derivative of the sigmoid function is unchanged (normal function). An good value for this constant might be 0.1.

Class org.joone.engine.WTALayer extends SimpleLayer implements Serializable

serialVersionUID: -941653911909171430L

Serialized Fields

LayerWidth

int LayerWidth

LayerHeight

int LayerHeight

LayerDepth

int LayerDepth
Package org.joone.engine.learning

Class org.joone.engine.learning.AbstractTeacherSynapse extends Synapse implements Serializable

serialVersionUID: -3501303723175798936L

Serialization Methods

readResolve

protected java.lang.Object readResolve()
Serialized Fields

desired

StreamInputSynapse desired
The stream from where to read the desired input.

net

NeuralNet net

lastErrorPatternReady

boolean lastErrorPatternReady
True if an error pattern is available

Class org.joone.engine.learning.ComparingSynapse extends java.lang.Object implements Serializable

serialVersionUID: -8893181016305737666L

Serialized Fields

theComparisonSynapse

ComparisonSynapse theComparisonSynapse

theLinearLayer

LinearLayer theLinearLayer

enabled

boolean enabled

outputFull

boolean outputFull

desired

StreamInputSynapse desired

monitor

Monitor monitor

name

java.lang.String name

Class org.joone.engine.learning.ComparisonSynapse extends Synapse implements Serializable

serialVersionUID: -1301682557631180066L

Serialized Fields

desired

StreamInputSynapse desired

Class org.joone.engine.learning.FahlmanTeacherSynapse extends TeacherSynapse implements Serializable

Serialized Fields

upperBit

double upperBit
The upperbit value (of the desired output), by default 1.

lowerBit

double lowerBit
The lowerbit value (of the desired output), by default 0.

lowerBitPercentage

double lowerBitPercentage
The percentage that is considered as a lowerbit, by default 0.4. In case of desired output bits {0, 1} any output within [0, 0.4] is considered a lower bit

upperBitPercentage

double upperBitPercentage
The percentage that is considered as a upperbit, by default 0.4. In case of desired output bits {0, 1} any output within [0.6, 1] is considered a upper bit

Class org.joone.engine.learning.TeacherSynapse extends AbstractTeacherSynapse implements Serializable

serialVersionUID: -1301682557631180066L

Class org.joone.engine.learning.TeachingSynapse extends java.lang.Object implements Serializable

serialVersionUID: -8893181016305737666L

Serialized Fields

theTeacherSynapse

AbstractTeacherSynapse theTeacherSynapse

theLinearLayer

LinearLayer theLinearLayer

enabled

boolean enabled

outputFull

boolean outputFull

theTeacherToUse

AbstractTeacherSynapse theTeacherToUse
The teacher to use. If null a normal TeacherSynapse will be used. The moment the teacher (theTeacherSynapse) is initialized is done the first time getTheTeacherSynapse() method is called. At that moment we are also able to set the monitor object.

desired

StreamInputSynapse desired

monitor

Monitor monitor

name

java.lang.String name
Package org.joone.engine.listeners

Class org.joone.engine.listeners.ConvergenceEvent extends java.util.EventObject implements Serializable

Class org.joone.engine.listeners.ConvergenceObserver extends MonitorPlugin implements Serializable

Serialized Fields

disableCurrentConvergence

boolean disableCurrentConvergence
The next flag indicates if the current convegence should be neglected. This is used in situations where convergence was reached, but the network continues running. If we would not neglect the current convergence event would continue to be generated. This flag is used to disable the event sfor the current convergence.

listeners

java.util.List<E> listeners
List of ConvergenceListeners.

Class org.joone.engine.listeners.DeltaBasedConvergenceObserver extends ConvergenceObserver implements Serializable

Serialized Fields

size

double size
Whenever each weight update value for some epochs / cycles is less than this size, a ConvergenceEvent will be generated.

cycles

int cycles
Sets the size of the cycles (sequence). Whenever the weight update values of a neural network are equal to or smaller than a certain value for this number of epochs, the network is considered as converged.

cycleCounter

int cycleCounter
Counter to check how many cycles the deltas where equal to or less than size.

net

NeuralNet net
The network holding the layers and synapses to be checked.

Class org.joone.engine.listeners.ErrorBasedConvergenceObserver extends ConvergenceObserver implements Serializable

Serialized Fields

percentage

double percentage
Whenever each training error in a sequence of errors decreases less than this percentage, a ConvergenceEvent will be generated.

cycles

int cycles
Sets the size of the cycles (sequence). Whenever the training error of a neural network is smaller than a certain percentage for this number of epochs, the network is considered as converged.

cycleCounter

int cycleCounter
Counter to check how many cycles the error decreases less than percentage.

lastError

double lastError
Variable to remember the previous error.

Class org.joone.engine.listeners.ErrorBasedTerminator extends MonitorPlugin implements Serializable

Serialized Fields

errorLevel

double errorLevel
The error level. Whenever the training error falls below this value the network should be stopped.

stoppedCycle

int stoppedCycle
The cycle the network was stopped.

stopRequested

boolean stopRequested
Has a stop request performed.

Package org.joone.engine.weights

Class org.joone.engine.weights.FanInBasedWeightInitializer extends java.lang.Object implements Serializable

Serialized Fields

lowerBound

double lowerBound
The lower bound.

upperBound

double upperBound
The upper bound.

sqrtFanIn

boolean sqrtFanIn
Flag indicating if we should use the square root of the fan-in (true), or should be use the normal fan-in (false) to determine the interval to init the weights with.

Class org.joone.engine.weights.RandomWeightInitializer extends java.lang.Object implements Serializable

serialVersionUID: 1547731234507850525L

Serialized Fields

upperBound

double upperBound
The upper boundery of the domain to initialize the weights with.

lowerBound

double lowerBound
The lower boundery of the domain to initialize the weights with.

Package org.joone.exception

Class org.joone.exception.JooneException extends java.lang.Exception implements Serializable

Serialized Fields

initialException

java.lang.Exception initialException

Class org.joone.exception.JooneRuntimeException extends java.lang.RuntimeException implements Serializable

Serialized Fields

initialException

java.lang.Throwable initialException

msg

java.lang.String msg
Package org.joone.io

Class org.joone.io.FileInputSynapse extends StreamInputSynapse implements Serializable

serialVersionUID: 7456627292136514416L

Serialization Methods

readObject

private void readObject(java.io.ObjectInputStream in)
                 throws java.io.IOException,
                        java.lang.ClassNotFoundException
Throws:
java.io.IOException
java.lang.ClassNotFoundException

writeObject

private void writeObject(java.io.ObjectOutputStream out)
                  throws java.io.IOException
Throws:
java.io.IOException
Serialized Fields

fileName

java.lang.String fileName
The name of the file to extract information from.

Class org.joone.io.FileOutputSynapse extends StreamOutputSynapse implements Serializable

serialVersionUID: 3194671306693862830L

Serialized Fields

FileName

java.lang.String FileName

append

boolean append

Class org.joone.io.ImageInputSynapse extends StreamInputSynapse implements Serializable

serialVersionUID: -1396287146739057882L

Serialized Fields

imageDirectory

java.io.File imageDirectory

theFileFilter

java.lang.String theFileFilter

FileNameList

java.util.Vector<E> FileNameList

MultiImages

java.awt.Image[] MultiImages

DesiredWidth

int DesiredWidth

DesiredHeight

int DesiredHeight

ColourMode

boolean ColourMode
If in Colour Mode RGB values are input seperately. Otherwise RGB values are averaged and input as grey scale.

Class org.joone.io.ImageOutputSynapse extends StreamOutputSynapse implements Serializable

serialVersionUID: 1287752494948170142L

Serialized Fields

JPG

int JPG

GIF

int GIF

PNG

int PNG

OutputDirectory

java.lang.String OutputDirectory

ImageFileType

int ImageFileType

Width

int Width

Height

int Height

ColourMode

boolean ColourMode

Class org.joone.io.InputConnector extends StreamInputSynapse implements Serializable

serialVersionUID: -3316265583123866079L

Serialized Fields

inputSynapse

StreamInputSynapse inputSynapse
The input stream this connector is connected to.

Class org.joone.io.InputSwitchSynapse extends StreamInputSynapse implements Serializable

serialVersionUID: 9025876263540714672L

Serialized Fields

inputs

java.util.Vector<E> inputs

activeSynapse

StreamInputSynapse activeSynapse

defaultSynapse

StreamInputSynapse defaultSynapse

name

java.lang.String name

mon

Monitor mon

outputDimension

int outputDimension

Class org.joone.io.JDBCInputSynapse extends StreamInputSynapse implements Serializable

serialVersionUID: -4642657913289986240L

Serialization Methods

readObject

private void readObject(java.io.ObjectInputStream in)
                 throws java.io.IOException,
                        java.lang.ClassNotFoundException
Reads this JDBCInputSynapse object into memory

Throws:
java.io.IOException - The Input Output Exception
java.lang.ClassNotFoundException - The class not found exception

writeObject

private void writeObject(java.io.ObjectOutputStream out)
                  throws java.io.IOException
Writes this JDBCSynapseInput object to the ObjectOutputStream out.

Throws:
java.io.IOException - The Input Output Exception if any
Serialized Fields

driverName

java.lang.String driverName
The name of the JDBC Database driver.

dbURL

java.lang.String dbURL
The URL of the database.

SQLQuery

java.lang.String SQLQuery
The database query to use in order to obtain the input data.

Class org.joone.io.JDBCOutputSynapse extends StreamOutputSynapse implements Serializable

serialVersionUID: 2176832390164459511L

Serialized Fields

driverName

java.lang.String driverName
The name of the JDBC Database driver.

dbURL

java.lang.String dbURL
The URL of the database.

SQLAmendment

java.lang.String SQLAmendment
The database query to use in order to obtain the input data.

Class org.joone.io.MemoryInputSynapse extends StreamInputSynapse implements Serializable

serialVersionUID: 2066217695979040186L

Serialized Fields

inputArray

double[][] inputArray

Class org.joone.io.MemoryOutputSynapse extends StreamOutputSynapse implements Serializable

serialVersionUID: 1342649629529016627L

Serialized Fields

patterns

Fifo patterns

zeroPattern

boolean zeroPattern

Class org.joone.io.MultipleInputSynapse extends InputSwitchSynapse implements Serializable

Serialized Fields

currentInput

int currentInput

currentPatt

int currentPatt

Class org.joone.io.StreamInputSynapse extends Synapse implements Serializable

serialVersionUID: -3316265583083866079L

Serialized Fields

firstRow

int firstRow

lastRow

int lastRow

firstCol

int firstCol

lastCol

int lastCol

advColumnsSel

java.lang.String advColumnsSel

buffered

boolean buffered
Flag indicating if the stream buffers the data of not.

decimalPoint

char decimalPoint

StepCounter

boolean StepCounter

plugIn

ConverterPlugIn plugIn

maxBufSize

int maxBufSize

plugInListeners

java.util.List<E> plugInListeners
List of plug-in listeners (often input connectors).

myFirstRow

int myFirstRow

myLastRow

int myLastRow

Class org.joone.io.StreamOutputSynapse extends Synapse implements Serializable

serialVersionUID: 7344684413113722785L

Serialized Fields

separator

char separator

buffered

boolean buffered

nextPlugIn

OutputConverterPlugIn nextPlugIn

Class org.joone.io.URLInputSynapse extends StreamInputSynapse implements Serializable

serialVersionUID: -1871585397469526608L

Serialization Methods

readObject

private void readObject(java.io.ObjectInputStream in)
                 throws java.io.IOException,
                        java.lang.ClassNotFoundException
Reads this URLInputSynapse object from the specified object stream.

Throws:
java.io.IOException
java.lang.ClassNotFoundException

writeObject

private void writeObject(java.io.ObjectOutputStream out)
                  throws java.io.IOException
Writes this URLInputSynapse object into the specified object stream.

Throws:
java.io.IOException
Serialized Fields

URL

java.lang.String URL
The string of the URL used to extract input data.

cURL

java.net.URL cURL
The actual URL used to extract input data.

Class org.joone.io.XLSInputSynapse extends StreamInputSynapse implements Serializable

serialVersionUID: 8625369117101456178L

Serialization Methods

readObject

private void readObject(java.io.ObjectInputStream in)
                 throws java.io.IOException,
                        java.lang.ClassNotFoundException
Reads this XLSInputSynpase object from the specified object stream.

Throws:
java.io.IOException
java.lang.ClassNotFoundException

writeObject

private void writeObject(java.io.ObjectOutputStream out)
                  throws java.io.IOException
Throws:
java.io.IOException
Serialized Fields

i_sheet_name

java.lang.String i_sheet_name
The name of the XLS worksheet from the XLS file to be used as input data.

i_sheet_index

int i_sheet_index
The work sheet index number.

file_chk

boolean file_chk
Flag to check the file name has been set.

fileName

java.lang.String fileName
The name of the XLS file to read data from.

Class org.joone.io.XLSOutputSynapse extends StreamOutputSynapse implements Serializable

serialVersionUID: -9167076940131905606L

Serialized Fields

FileName

java.lang.String FileName
The name of the XLS File to write network output data to.

o_sheet_index

int o_sheet_index
The worksheet index number.

row_no

int row_no
Internal row count.

startCol

int startCol
Starting position when the first pattern being written.

startRow

int startRow

o_sheet_name

java.lang.String o_sheet_name
The name of the XLS sheet within the XLS File that will be used to write data to.

Class org.joone.io.YahooFinanceInputSynapse extends StreamInputSynapse implements Serializable

serialVersionUID: 1301769209320717393L

Serialization Methods

readObject

private void readObject(java.io.ObjectInputStream in)
                 throws java.io.IOException,
                        java.lang.ClassNotFoundException
Reads this YahooFinanceInputSynapse object into memory from the specified object stream.

Throws:
java.io.IOException - The Input Output Exception
java.lang.ClassNotFoundException - The class not found exception

writeObject

private void writeObject(java.io.ObjectOutputStream out)
                  throws java.io.IOException
Writes this YahooFInanceSynapseInput object to the ObjectOutputStream out.

Throws:
java.io.IOException - The Input Output Exception if any
Serialized Fields

months

java.lang.String[] months

frequency

java.lang.String[] frequency

freq_conv

java.lang.String[] freq_conv

Symbol

java.lang.String Symbol

date_formater

java.text.DateFormat date_formater

CalendarStart

java.util.Calendar CalendarStart

CalendarEnd

java.util.Calendar CalendarEnd

DateStart

java.lang.String DateStart

DateEnd

java.lang.String DateEnd

Period

java.lang.String Period

StockData

java.util.Vector<E>[] StockData

StockDates

java.util.Vector<E> StockDates

ColumnNames

java.lang.String[] ColumnNames
Package org.joone.net

Class org.joone.net.NestedNeuralLayer extends Layer implements Serializable

serialVersionUID: -3697306754884303651L

Serialization Methods

readObject

private void readObject(java.io.ObjectInputStream in)
                 throws java.io.IOException,
                        java.lang.ClassNotFoundException
Throws:
java.io.IOException
java.lang.ClassNotFoundException
Serialized Fields

sNeuralNet

java.lang.String sNeuralNet

NestedNeuralNet

NeuralNet NestedNeuralNet

lin

LinearLayer lin

Class org.joone.net.NeuralNet extends java.lang.Object implements Serializable

serialVersionUID: 8351124226081783962L

Serialization Methods

readObject

private void readObject(java.io.ObjectInputStream in)
                 throws java.io.IOException,
                        java.lang.ClassNotFoundException
Throws:
java.io.IOException
java.lang.ClassNotFoundException
Serialized Fields

layers

java.util.Vector<E> layers

NetName

java.lang.String NetName

mon

Monitor mon

inputLayer

Layer inputLayer

outputLayer

Layer outputLayer

teacher

ComparingElement teacher

listeners

java.util.Vector<E> listeners

macroPlugin

MacroInterface macroPlugin

scriptingEnabled

boolean scriptingEnabled

descriptor

NeuralNetAttributes descriptor

params

java.util.Hashtable<K,V> params

orderedLayers

Layer[] orderedLayers

stopFastRun

boolean stopFastRun
Implementation code for the single-thread version of Joone

patternNumber

int patternNumber
The number of the current pattern being chased through the network

Class org.joone.net.NeuralNetAttributes extends java.lang.Object implements Serializable

serialVersionUID: -3122881040378874490L

Serialized Fields

validationError

double validationError

trainingError

double trainingError

neuralNetName

java.lang.String neuralNetName

lastEpoch

int lastEpoch

Class org.joone.net.NeuralValidationEvent extends java.util.EventObject implements Serializable

Package org.joone.script

Class org.joone.script.JooneMacro extends java.lang.Object implements Serializable

serialVersionUID: 6361561451436197429L

Serialized Fields

text

java.lang.String text

event

boolean event

name

java.lang.String name

Class org.joone.script.MacroManager extends java.lang.Object implements Serializable

serialVersionUID: 2855350620727616763L

Serialized Fields

macros

java.util.Hashtable<K,V> macros
Package org.joone.structure

Class org.joone.structure.Nakayama extends java.lang.Object implements Serializable

Serialized Fields

layers

java.util.List<E> layers
List with layers to be optimized (layers should be on the same (layer) level).

isRunning

boolean isRunning
Flag to remember if the network was running when the request come to optimize the network. If so we will re-start it after optimization.

net

NeuralNet net
The net to optimize.

clone

NeuralNet clone
The clone of a network to collect information (outputs of neurons after a pattern has been forwarded through the network).

epsilon

double epsilon
The threshold to decide if the neuron should be deleted or not, i.e. if Cj = Ij / max-j{Ij} * Vj / max-j{Vj} * Rjmin <= epsilon.

listeners

java.util.List<E> listeners
The original listeners of the neural network that will be temporarely removed. We will add them again after optimization when we restart the network.

outputsAfterPattern

java.util.List<E> outputsAfterPattern
Holds the outputs of the neurons after each pattern [pattern->layer->neuron].

information

java.util.List<E> information
Holds the information for every neuron (excluding input and output neurons) to its output layers [layer->neuron].

infoMax

double infoMax
The maximum value of the information from a neuron to its output layer.

averageOutputs

java.util.List<E> averageOutputs
Holds the average output over all patterns for every neuron (excluding input and output neurons) [layer->neuron].

variance

java.util.List<E> variance
Holds the variance for every neuron.

varianceMax

double varianceMax
The maximum variance value [layer->neuron].

gamma

java.util.List<E> gamma
Holds the gamma value for every neuron [layer->neuron->layer->neuron].

optimized

boolean optimized
Flag indicating if the network is optimized, i.e. if neurons are removed.

Class org.joone.structure.NetworkLayer extends Layer implements Serializable

Serialized Fields

nodeList

java.util.List<E> nodeList
List of all nodes

inputNodes

java.util.List<E> inputNodes
List of input nodes

outputNodes

java.util.List<E> outputNodes
List of output nodes

contextNodes

java.util.List<E> contextNodes
List of all context nodes

orderedNodes

java.util.List<E> orderedNodes
Ordered nodes

fixed

boolean fixed
True if context node are in fixed state - useful when simulating

patternCount

int patternCount
Pattern counter

inputSynapse

DirectSynapse inputSynapse
Synapse used to interrogate network

Class org.joone.structure.NetworkViewer extends javax.swing.JPanel implements Serializable

Serialized Fields

network

NetworkLayer network
The network we are plotting here

horizontalSpacingLabel

javax.swing.JLabel horizontalSpacingLabel

horizontalSpacingSpinner

javax.swing.JSpinner horizontalSpacingSpinner

networkViewerTabbedPane

javax.swing.JTabbedPane networkViewerTabbedPane

nodeViewerCommandPanel

javax.swing.JPanel nodeViewerCommandPanel

nodeViewerPanel

javax.swing.JPanel nodeViewerPanel

nodeViewerScrollPane

javax.swing.JScrollPane nodeViewerScrollPane

plotButton

javax.swing.JButton plotButton

treeViewerScrollPane

javax.swing.JScrollPane treeViewerScrollPane

treeViewerTree

javax.swing.JTree treeViewerTree

verticalSpacingLabel

javax.swing.JLabel verticalSpacingLabel

verticalSpacingSpinner

javax.swing.JSpinner verticalSpacingSpinner
Package org.joone.util

Class org.joone.util.AbstractConverterPlugIn extends java.lang.Object implements Serializable

serialVersionUID: 5698511686417862414L

Serialization Methods

readObject

private void readObject(java.io.ObjectInputStream in)
                 throws java.io.IOException,
                        java.lang.ClassNotFoundException
Throws:
java.io.IOException
java.lang.ClassNotFoundException
Serialized Fields

nextPlugIn

AbstractConverterPlugIn nextPlugIn
The next plugin in this series of cascading plugins.

name

java.lang.String name
The name of this plug-in object.

connected

boolean connected
This flag indicates if this plug-in is connected. Whenever a plug in is connected it cannot be connected / added to another input stream / plug-in.

pluginListeners

java.util.Vector<E> pluginListeners
A vector of objects that are listening to this object for plug-in (data changed) events.

AdvancedSerieSelector

java.lang.String AdvancedSerieSelector
The AdvancedSerieSelector instructs this plug-in what serie/columns it should process. The format of this specification is a common separated list of values and ranges. E.g '1,2,5,7' will instruct the converter to convert serie 1 and 2 and 5 and 7. A range can also be used e.g '2,4,5-8,9' will instruct the converter to process serie 2 and 4 and 5 and 6 and 7 and 8 and 9. A range is specifed using a '-' character with the number of the serie on either side.

Note NO negative numbers can be used in the AdvancedSerieSelector.

Class org.joone.util.CenterOnZeroPlugIn extends ConverterPlugIn implements Serializable

serialVersionUID: 8581778588210471901L

Class org.joone.util.ColumnSelectorPlugIn extends ConverterPlugIn implements Serializable

Class org.joone.util.ConverterPlugIn extends AbstractConverterPlugIn implements Serializable

serialVersionUID: 1698511686417967414L

Serialization Methods

readObject

private void readObject(java.io.ObjectInputStream in)
                 throws java.io.IOException,
                        java.lang.ClassNotFoundException
Throws:
java.io.IOException
java.lang.ClassNotFoundException
Serialized Fields

applyEveryCycle

boolean applyEveryCycle
Flag indicating if every cycle the data should be preprocesed.

Class org.joone.util.DeltaNormPlugIn extends ConverterPlugIn implements Serializable

serialVersionUID: 1698511686417955514L

Class org.joone.util.DynamicAnnealing extends MonitorPlugin implements Serializable

serialVersionUID: -5494365758818313237L

Serialized Fields

lastError

double lastError

step

double step

Class org.joone.util.GroovyMacroPlugin extends MonitorPlugin implements Serializable

serialVersionUID: -4123807261022916301L

Serialized Fields

macroManager

MacroManager macroManager

Class org.joone.util.InputPluginEvent extends PlugInEvent implements Serializable

Class org.joone.util.LearningSwitch extends InputSwitchSynapse implements Serializable

serialVersionUID: -2339515807277374407L

Serialized Fields

trainingSet

StreamInputSynapse trainingSet

validationSet

StreamInputSynapse validationSet

validation

boolean validation

validationPatterns

int validationPatterns
Not more used. They are here only to maintain the compatibility with the old serialized classes. Use instead the patterns/validationPatterns parameters of the Monitor object.

trainingPatterns

int trainingPatterns

Class org.joone.util.LinearAnnealing extends MonitorPlugin implements Serializable

serialVersionUID: -3786770944656325519L

Serialized Fields

learningRateInitial

double learningRateInitial

learningRateFinal

double learningRateFinal

momentumInitial

double momentumInitial

momentumFinal

double momentumFinal

Class org.joone.util.LogarithmicPlugIn extends ConverterPlugIn implements Serializable

serialVersionUID: 4662839350631574552L

Class org.joone.util.MacroPlugin extends MonitorPlugin implements Serializable

serialVersionUID: -4867807261022916301L

Serialized Fields

macroManager

MacroManager macroManager

Class org.joone.util.MinMaxExtractorPlugIn extends ConverterPlugIn implements Serializable

serialVersionUID: 6835532403570359948L

Serialized Fields

MIN_VALUE

double MIN_VALUE

MAX_VALUE

double MAX_VALUE

MID_VALUE

double MID_VALUE

minPerc

double minPerc

Class org.joone.util.MonitorPlugin extends java.lang.Object implements Serializable

serialVersionUID: 951079164859904152L

Serialized Fields

name

java.lang.String name

rate

int rate

neuralNet

NeuralNet neuralNet

Class org.joone.util.MovingAveragePlugIn extends ConverterPlugIn implements Serializable

serialVersionUID: -5679399800426091523L

Serialized Fields

AdvancedMovAvgSpec

java.lang.String AdvancedMovAvgSpec

Class org.joone.util.NormalizerPlugIn extends ConverterPlugIn implements Serializable

serialVersionUID: 4662839350631576461L

Serialized Fields

min

double min

max

double max

datamin

double datamin

datamax

double datamax

Class org.joone.util.OutputConverterPlugIn extends AbstractConverterPlugIn implements Serializable

serialVersionUID: 1698511686417967414L

Serialization Methods

readObject

private void readObject(java.io.ObjectInputStream in)
                 throws java.io.IOException,
                        java.lang.ClassNotFoundException
Throws:
java.io.IOException
java.lang.ClassNotFoundException

Class org.joone.util.OutputPluginEvent extends PlugInEvent implements Serializable

Class org.joone.util.PlugInEvent extends java.util.EventObject implements Serializable

Class org.joone.util.RbfRandomCenterSelector extends ConverterPlugIn implements Serializable

Serialized Fields

theRbfGaussianLayer

RbfGaussianLayer theRbfGaussianLayer
The RBF layer (to get the number of neurons).

thePatterns

java.util.Vector<E> thePatterns
Save the input vectors.

Class org.joone.util.ShufflePlugin extends ConverterPlugIn implements Serializable

serialVersionUID: 7118539833128121178L

Class org.joone.util.SnapshotPlugin extends MonitorPlugin implements Serializable

serialVersionUID: -4796324031568433167L

Class org.joone.util.SnapshotRecorder extends SnapshotPlugin implements Serializable

serialVersionUID: -8151866025667526018L

Serialized Fields

filename

java.lang.String filename

format

java.lang.String format

Class org.joone.util.ToBinaryPlugin extends ConverterPlugIn implements Serializable

Serialized Fields

theConvertedSeries

java.util.List<E> theConvertedSeries
The sizes of the (binary) arrays of converted series. This way we are able to find the correct position of serie we have to convert related taking into account any previous converted series.

upperBit

double upperBit
The value for the upper bit.

lowerBit

double lowerBit
The value for the lower bit.

Class org.joone.util.UnNormalizerOutputPlugIn extends OutputConverterPlugIn implements Serializable

serialVersionUID: 5322361123972428588L

Serialized Fields

min

double min
The Out Data Min property.

max

double max
The Out Data Max property.

datamax

double datamax
The In Data Max property.

datamin

double datamin
The In Data Min property.

Class org.joone.util.WeightWriterPlugin extends MonitorPlugin implements Serializable

Serialized Fields

weightWriter

java.io.PrintWriter weightWriter
The writer to which the weights are dumped

nodesAndWeights

NodesAndWeights nodesAndWeights
Nodes and weights used to obtain information

Overview  Package  Class  Use  Tree  Deprecated  Index  Help 
 PREV   NEXT FRAMES    NO FRAMES    
Submit Feedback to pmarrone@users.sourceforge.net