Package | Description |
---|---|
org.joone.engine | |
org.joone.engine.listeners | |
org.joone.net | |
org.joone.script | |
org.joone.structure | |
org.joone.util |
Modifier and Type | Class and Description |
---|---|
class |
EKFFFNLearnerPlugin
A plugin listener that implements the EKFFFN learner used
to train feed forward neural networks.
|
class |
EKFRNNLearnerPlugin
A plugin listener that implements the EKF learner, based on
"Some observations on the use of the extended Kalman filter
as a recurrent network learning algorithm" by Williams (1992)
in order to train a network.
|
class |
GaussianLayer
This layer implements the Gaussian Neighborhood SOM strategy.
|
class |
KohonenSynapse
This is an unsupervised Kohonen Synapse which is a Self Organising Map.
|
class |
NeuralNetAdapter |
class |
RTRLLearnerFactory
A RTRL implementation.
|
class |
RTRLLearnerPlugin
A plugin listener that applies the RTRL algorithm to a network.
|
Modifier and Type | Method and Description |
---|---|
void |
Monitor.addNeuralNetListener(NeuralNetListener l)
adds a neural net event listener the Monitor
|
void |
Monitor.addNeuralNetListener(NeuralNetListener l,
boolean removable)
adds a neural net event listener to the Monitor
|
protected void |
Layer.removeListener(NeuralNetListener listener) |
void |
Monitor.removeNeuralNetListener(NeuralNetListener l)
Removes a listener
|
Modifier and Type | Class and Description |
---|---|
class |
ConvergenceObserver
Abstract class for all convergence observer.
|
class |
DeltaBasedConvergenceObserver
This observer observes if the network has convergenced based on the size of the
weight updates (deltas).
|
class |
ErrorBasedConvergenceObserver
This observer observes if the network has convergenced based on the
sequence of training errors.
|
class |
ErrorBasedTerminator
Stops a network whenever the training error of the network falls below a
certain value.
|
Modifier and Type | Class and Description |
---|---|
class |
NeuralNet
This object represents a container of a neural network,
giving to the developer the possibility to manage a
neural network as a whole.
|
class |
NeuralNetValidator
This class is useful to validate a neural network.
|
Modifier and Type | Method and Description |
---|---|
void |
NeuralNet.addNeuralNetListener(NeuralNetListener listener) |
void |
NeuralNet.removeNeuralNetListener(NeuralNetListener listener) |
Modifier and Type | Interface and Description |
---|---|
interface |
MacroInterface |
Modifier and Type | Class and Description |
---|---|
class |
Nakayama
This class performs the method of optimizing activation functions as described
in:
|
Modifier and Type | Class and Description |
---|---|
class |
DynamicAnnealing
This plugin controls the change of the learning rate based on the difference
between the last two global error (E) values:
if E(t) > E(t-1) then LR = LR * (1 - step/100)
Note: step/100 because step is inserted as a % value from the user
if E(t) <= E(t-1) then LR is unchanged
|
class |
GroovyMacroPlugin |
class |
LinearAnnealing
This plugin changes linearly the values of the learning rate and of the momentum parameters.
|
class |
MacroPlugin |
class |
MonitorPlugin
This class represents a generic listener of the net's events.
|
class |
NeuralNetRunner |
class |
SnapshotPlugin
A SnapshotPlugin manages the generation of snapshots of the network's state
at regular intervals during its activity.
|
class |
SnapshotRecorder
A SnapshotRecorder serves to create and record snapshots of a neural network
in a file as a serie of serialized objects graphs.
|
class |
WeightWriterPlugin
A class that can be used to write weight values during training to
a file in gnuplot format.
|
Submit Feedback to pmarrone@users.sourceforge.net