Package com.anji.integration

com.anji.integration contains classes that provide the glue between the various components of the system, including artificial neural net implementations, JGAP, NEAT, Log4J, and persistence and presentation layers.

See:
Description

Interface Summary

Activator	Although any object taking and returning double arrays can implement this, it is meant to be a simple neural net interface where the arrays are input and output activation patterns.
Transcriber	To "transcribe" is to construct a phenotype from a genotype.

Class Summary

ActivatorTranscriber	Factory interface to abstract construction of neural network objects.
AnjiActivator	Simple neural network implementation of Activator interface.
AnjiNetTranscriber	The purpose of this class is to construct a neural net object (AnjiNet) from a chromosome.
ChromosomeToNetworkXml	Converts chromosome to activator, which it then uses to write network XML according to NEVT data model.
ConsoleLogEventListener	Writes log events to stdout.
ErrorFunction
ErrorRateCounter
Evaluator
ExponentialTargetFitnessFunction	Fitness function where error is exponential; i.e., as the error gets closer to 0, the fitness increases exponentially, at a greater rate than simply squaring the error.
Generation	Converts generation data between Genotype and XML.
LogEventListener	Writes log events to log4j framework.
PersistenceEventListener	Writes genetic algorithm data, including chromosomes and run, to persistent storage.
PresentationEventListener	Stores presentation XML data where it can be displayed, likely in a web browser.
RandomFitnessFunction	Assigns random fitness for each chromosome.
SimpleSelector	Selects chromosomes based directly on fitness value, as opposed to a statistical probability.
TargetFitnessFunction	Determines fitness based on how close Activator output is to a target.
XmlPersistableAllele
XmlPersistableChromosome
XmlPersistableRun	Converts run data to XML presentation format.

Exception Summary

AnjiRequiredException	Thrown if ANJI-specific JGAP objects receive a non-ANJI configuration.
TranscriberException

Package com.anji.integration Description

com.anji.integration contains classes that provide the glue between the various components of the system, including artificial neural net implementations, JGAP, NEAT, Log4J, and persistence and presentation layers. com.anji.neat also contains some integration code between NEAT and JGAP.