org.encog.neural.networks.training.propagation
Interface PropagationMethod

All Known Implementing Classes:

BackpropagationMethod, ManhattanPropagationMethod, ResilientPropagationMethod

public interface PropagationMethod

Defines the specifics to one of the propagation methods. The individual ways that each of the propagation methods uses to modify the weight and] threshold matrix are defined here.

Author:

jheaton

Method Summary

void	calculateError(NeuralOutputHolder output, PropagationLevel fromLevel, PropagationLevel toLevel) Calculate the error between these two levels.
void	init(Propagation propagation) Init with the specified propagation object.
void	learn() Apply the accumulated deltas and learn.

Method Detail

calculateError

void calculateError(NeuralOutputHolder output,
                    PropagationLevel fromLevel,
                    PropagationLevel toLevel)

Calculate the error between these two levels.

Parameters:

output - The output to the "to level".

fromLevel - The from level.

toLevel - The target level.

init

void init(Propagation propagation)

Init with the specified propagation object.

Parameters:

propagation - The propagation object that this method will be used with.

learn

void learn()

Apply the accumulated deltas and learn.