clarion.system
Interface InterfaceTrainable

All Known Subinterfaces:
InterfaceRuntimeTrainable, InterfaceUsesQLearning
All Known Implementing Classes:
AbstractNeuralNet, AbstractRuntimeTrainableBPNet, AbstractRuntimeTrainableImplicitModule, AbstractTrainableImplicitModule, BPNet, QBPNet, SimplifiedQBPNet, TableLookup

public interface InterfaceTrainable

This interface is implemented by classes that are trainable within CLARION.

Usage:

Any class that implements this interface will be capable of being trained "off-line" at a minimum. If you wish to build a class that can be trained at runtime, implement the InterfaceRuntimeTrainable interface instead.

Known Subinterfaces:

Known classes implementing this interface:

Version:
6.0.4
Author:
Nick Wilson

Method Summary
 void backwardPass()
          Performs a backward pass, "learning" based on the desired output.
 double getSumSqErrors()
          Gets the sum of squared errors.
 void setDesiredOutput(AbstractOutputChunk DesiredOut)
          Sets a specific desired output node given the specified chunk.
 void setDesiredOutput(java.util.Collection<? extends AbstractOutputChunk> DesiredOut)
          Sets the desired output given the collection specified.
 

Method Detail

setDesiredOutput

void setDesiredOutput(java.util.Collection<? extends AbstractOutputChunk> DesiredOut)
Sets the desired output given the collection specified.

Parameters:
DesiredOut - The desired output to be set.

setDesiredOutput

void setDesiredOutput(AbstractOutputChunk DesiredOut)
Sets a specific desired output node given the specified chunk.

Parameters:
DesiredOut - The desired output chunk to be set.

backwardPass

void backwardPass()
Performs a backward pass, "learning" based on the desired output.


getSumSqErrors

double getSumSqErrors()
Gets the sum of squared errors.

Returns:
The sum of squared errors.