Class NetworkInvoker

java.lang.Object

org.ojalgo.ann.WrappedANN

org.ojalgo.ann.NetworkInvoker

All Implemented Interfaces:

Supplier<ArtificialNeuralNetwork>, BasicFunction, BasicFunction.PlainUnary<Access1D<Double>, MatrixStore<Double>>

public class NetworkInvoker
extends WrappedANN
implements BasicFunction.PlainUnary<Access1D<Double>, MatrixStore<Double>>

Nested Class Summary

Nested classes/interfaces inherited from interface BasicFunction

BasicFunction.Differentiable<N,F>, BasicFunction.Integratable<N,F>, BasicFunction.PlainUnary<T,R>

Constructor Summary

Constructors

Constructor	Description
NetworkInvoker(ArtificialNeuralNetwork network, int batchSize)

Method Summary

Modifier and Type	Method	Description
MatrixStore<Double>	invoke(Access1D<Double> input)	The input argument is typed as Access1D which essentially means it can be anything.

Methods inherited from class WrappedANN

adjust, depth, equals, get, getActivator, getBatchSize, getBias, getInput, getInput, getOutput, getOutput, getOutputActivator, getWeight, getWeights, hashCode, invoke, newInputBatch, newOutputBatch, randomise, setActivator, setBias, setWeight, structure

Methods inherited from class Object

clone, finalize, getClass, notify, notifyAll, toString, wait, wait, wait

Constructor Details

NetworkInvoker

NetworkInvoker(ArtificialNeuralNetwork network,
 int batchSize)

Method Details

invoke

public MatrixStore<Double> invoke(Access1D<Double> input)

The input argument is typed as Access1D which essentially means it can be anything. If the batch size is anything other than 1 this needs to be a 2D structure with the number of rows matching the batch size. The return type is a MatrixStore where the number of rows match the batch size and the number of columns match the number of output nodes. A DataBatch may help when creating the batches or you simply create any 2D data structure and fill the rows.

Specified by:

invoke in interface BasicFunction.PlainUnary<Access1D<Double>, MatrixStore<Double>>

See Also:

• BasicFunction.PlainUnary.invoke(java.lang.Object)