public static enum ArtificialNeuralNetwork.Error extends Enum<ArtificialNeuralNetwork.Error> implements PrimitiveFunction.Binary
BinaryFunction.FixedFirst<N extends Number>, BinaryFunction.FixedSecond<N extends Number>
BasicFunction.Differentiable<N extends Number,F extends BasicFunction>, BasicFunction.Integratable<N extends Number,F extends BasicFunction>, BasicFunction.PlainUnary<T,R>
Enum Constant and Description |
---|
CROSS_ENTROPY
Currently this can only be used in in combination with
ArtificialNeuralNetwork.Activator.SOFTMAX in the final
layer. |
HALF_SQUARED_DIFFERENCE |
Modifier and Type | Method and Description |
---|---|
double |
invoke(Access1D<?> target,
Access1D<?> current) |
double |
invoke(double target,
double current) |
static ArtificialNeuralNetwork.Error |
valueOf(String name)
Returns the enum constant of this type with the specified name.
|
static ArtificialNeuralNetwork.Error[] |
values()
Returns an array containing the constants of this enum type, in
the order they are declared.
|
clone, compareTo, equals, finalize, getDeclaringClass, hashCode, name, ordinal, toString, valueOf
invoke
andThen, apply, applyAsDouble, by, by, first, first, second, second
maxBy, minBy
andThen
public static final ArtificialNeuralNetwork.Error CROSS_ENTROPY
ArtificialNeuralNetwork.Activator.SOFTMAX
in the final
layer. All other usage will give incorrect network training.public static final ArtificialNeuralNetwork.Error HALF_SQUARED_DIFFERENCE
public static ArtificialNeuralNetwork.Error[] values()
for (ArtificialNeuralNetwork.Error c : ArtificialNeuralNetwork.Error.values()) System.out.println(c);
public static ArtificialNeuralNetwork.Error valueOf(String name)
name
- the name of the enum constant to be returned.IllegalArgumentException
- if this enum type has no constant with the specified nameNullPointerException
- if the argument is nullpublic double invoke(double target, double current)
invoke
in interface BinaryFunction<Double>
Copyright © 2019 Optimatika. All rights reserved.