com.github.mrdimosthenis.synapses

Class Fun

java.lang.Object

com.github.mrdimosthenis.synapses.Fun

public class Fun
extends java.lang.Object

The activation functions a neuron can have.

They can be used in the arguments of neural network's constructor.

Field Summary

Fields

Modifier and Type	Field and Description
static Fun	IDENTITY Identity is a linear function where the output is equal to the input.
static Fun	LEAKY_RE_LU LeakyReLU gives a small proportion of x if x is negative and x otherwise.
static Fun	SIGMOID Sigmoid takes any real value as input and outputs values in the range of 0 to 1.
static Fun	TANH Tanh is similar to Sigmoid, but outputs values in the range of -1 and 1.

Method Summary

Methods inherited from class java.lang.Object

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

Field Detail

SIGMOID

public static final Fun SIGMOID

Sigmoid takes any real value as input and outputs values in the range of 0 to 1.

 
 x -> 1.0 / (1.0 + Math.exp(-x))
 
 

IDENTITY

public static final Fun IDENTITY

Identity is a linear function where the output is equal to the input.

 
 x -> x
 
 

TANH

public static final Fun TANH

Tanh is similar to Sigmoid, but outputs values in the range of -1 and 1.

 
 x -> Math.tanh(x)
 
 

LEAKY_RE_LU

public static final Fun LEAKY_RE_LU

LeakyReLU gives a small proportion of x if x is negative and x otherwise.

 
 x -> x < 0.0 ? 0.01 * x : x