NAME

AI::ActivationFunctions - Activation functions for neural networks in Perl

VERSION

Version 0.01

ABSTRACT

Activation functions for neural networks in Perl

SYNOPSIS

use AI::ActivationFunctions qw(relu prelu sigmoid);

my $result = relu(-5);  # returns 0
my $prelu_result = prelu(-2, 0.1);  # returns -0.2

# Array version works too
my $array_result = relu([-2, -1, 0, 1, 2]);  # returns [0, 0, 0, 1, 2]

DESCRIPTION

This module provides various activation functions commonly used in neural networks and machine learning. It includes basic functions like ReLU and sigmoid, as well as advanced functions like GELU and Swish.

FUNCTIONS

Basic Functions

• relu($input)

  Rectified Linear Unit. Returns max(0, $input).

• prelu($input, $alpha=0.01)

  Parametric ReLU. Returns $input if $input > 0, else $alpha * $input.

• leaky_relu($input)

  Leaky ReLU with alpha=0.01.

• sigmoid($input)

  Sigmoid function: 1 / (1 + exp(-$input)).

• tanh($input)

  Hyperbolic tangent function.

• softmax(\@array)

  Softmax function for probability distributions.

Advanced Functions

• elu($input, $alpha=1.0)

  Exponential Linear Unit.

• swish($input)

  Swish activation function.

• gelu($input)

  Gaussian Error Linear Unit (used in transformers like BERT, GPT).

Derivatives

• relu_derivative($input)

  Derivative of ReLU for backpropagation.

• sigmoid_derivative($input)

  Derivative of sigmoid for backpropagation.

EXPORT

By default nothing is exported. You can export specific functions:

use AI::ActivationFunctions qw(relu prelu);  # specific functions
use AI::ActivationFunctions qw(:basic);      # basic functions
use AI::ActivationFunctions qw(:all);        # all functions

SEE ALSO

• PDL - Perl Data Language for numerical computing

• AI::TensorFlow - Perl interface to TensorFlow

• AI::MXNet - Perl interface to Apache MXNet

AUTHOR

Your Name <your.email@example.com>

LICENSE

This library is free software; you can redistribute it and/or modify it under the same terms as Perl itself.