Class ELU

java.lang.Object

smile.deep.activation.ActivationFunction

smile.deep.activation.ELU

All Implemented Interfaces:

Serializable, Function<Tensor,Tensor>, Layer

public class ELU
extends ActivationFunction

Exponential Linear Unit (ELU) activation function.

ELU is defined as:

• x if x > 0
• alpha * (exp(x) - 1) if x <= 0

Unlike ReLU, ELU produces negative outputs, pushing the mean activation closer to zero, which can speed up training.

See Also:

• Fast and Accurate Deep Network Learning by Exponential Linear Units (ELUs)
• Serialized Form

Constructor Summary

Constructors

Constructor	Description
ELU()	Constructor with default alpha = 1.0.
ELU(double alpha, boolean inplace)	Constructor.

Method Summary

Modifier and Type	Method	Description
double	alpha()	Returns the alpha value.
Tensor	forward(Tensor input)	Forward propagation (or forward pass) through the layer.

Methods inherited from class ActivationFunction

asTorch, isInplace, name

Methods inherited from class Object

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

Methods inherited from interface Function

andThen, compose

Methods inherited from interface Layer

apply, to, to

Constructor Details

ELU

public ELU()

Constructor with default alpha = 1.0.

ELU

public ELU(double alpha,
 boolean inplace)

Constructor.

Parameters:

alpha - the alpha value for the ELU formulation. Must be non-negative.

inplace - true if the operation executes in-place.

Method Details

alpha

public double alpha()

Returns the alpha value.

Returns:

the alpha value.

forward

public Tensor forward(Tensor input)

Description copied from interface: Layer

Forward propagation (or forward pass) through the layer.

Parameters:

input - the input tensor.

Returns:

the output tensor.