nuroko.core

Class Components

java.lang.Object

nuroko.core.Components

public final class Components
extends Object

Constructor Summary

Constructors

Constructor and Description
Components()

Method Summary

Modifier and Type	Method and Description
static ALayerStack	asLayerStack(IComponent comp)
static IComponent	bias(int length)
static LossFunction	defaultLossFunction(mikera.vectorz.Op op) Returns a default loss function for a specific operator
static Dropout	dropout(int length)
static Dropout	dropout(int length, double dropoutRate)
static Identity	identity(int length)
static Join	join(IComponent... components)
static Join	join(List<? extends IComponent> components)
static NeuralNet	neuralLayer(int inputLength, int outputLength, mikera.vectorz.Op op)
static NeuralNet	neuralLayer(int inputLength, int outputLength, mikera.vectorz.Op op, boolean fullyConnected)
static NeuralNet	neuralLayer(int inputLength, int outputLength, mikera.vectorz.Op op, int maxLinks)
static Operator	offset(int length, double offset)
static Operator	operator(mikera.vectorz.Op op, int length)
static Stack	stack(IComponent... components)
static Stack	stack(List<? extends IComponent> components)
static AWeightLayer	weightLayer(int inputLength, int outputLength, int maxLinks)

Methods inherited from class java.lang.Object

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

Constructor Detail

Components

public Components()

Method Detail

stack

public static Stack stack(List<? extends IComponent> components)

stack

public static Stack stack(IComponent... components)

join

public static Join join(List<? extends IComponent> components)

join

public static Join join(IComponent... components)

operator

public static Operator operator(mikera.vectorz.Op op,
                                int length)

neuralLayer

public static NeuralNet neuralLayer(int inputLength,
                                    int outputLength,
                                    mikera.vectorz.Op op)

neuralLayer

public static NeuralNet neuralLayer(int inputLength,
                                    int outputLength,
                                    mikera.vectorz.Op op,
                                    int maxLinks)

dropout

public static Dropout dropout(int length)

dropout

public static Dropout dropout(int length,
                              double dropoutRate)

weightLayer

public static AWeightLayer weightLayer(int inputLength,
                                       int outputLength,
                                       int maxLinks)

neuralLayer

public static NeuralNet neuralLayer(int inputLength,
                                    int outputLength,
                                    mikera.vectorz.Op op,
                                    boolean fullyConnected)

asLayerStack

public static ALayerStack asLayerStack(IComponent comp)

defaultLossFunction

public static LossFunction defaultLossFunction(mikera.vectorz.Op op)

Returns a default loss function for a specific operator

Parameters:

topOp -

Returns:

identity

public static Identity identity(int length)

offset

public static Operator offset(int length,
                              double offset)

bias

public static IComponent bias(int length)