view release on metacpan or  search on metacpan

lib/Algorithm/FeatureSelection.pm  view on Meta::CPAN

package Algorithm::FeatureSelection;
use strict;
use warnings;
use List::Util qw(sum);

our $VERSION = '0.02';

sub new {
    my $class = shift;
    my $self = bless {@_}, $class;
    return $self;
}

sub pmi {
    my $self = shift;
    $self->pairewise_mutual_information(@_);
}

sub ig {
    my $self = shift;
    $self->information_gain(@_);
}

sub igr {
    my $self = shift;
    $self->information_gain_ratio(@_);
}

sub pairwise_mutual_information {
    my $self     = shift;
    my $features = shift;

    ## -----------------------------------------------------------------
    ##
    ## The argument is expected as below.
    ##
    ## $features = {
    ##     feature_1 => {
    ##         class_a => 10,
    ##         class_b => 2,
    ##     },
    ##     feature_2 => {
    ##         class_b => 11,
    ##         class_d => 32
    ##     },
    ##           .
    ##           .
    ##           .
    ## };
    ##
    ## -----------------------------------------------------------------
    ##
    ## Pairewise Mutual Information
    ##
    ## PMI(w, c) = log ( P( Xw = 1, C = c ) / P( Xw=1 )P( C=c ) )
    ##
    ##    c.f.  w = feature
    ##          c = Class
    ##
    ## -----------------------------------------------------------------

    my $feature_count;
    my $class_count;
    my $co_occur_count;
    my $all_features_num;
    while ( my ( $feature, $ref ) = each %$features ) {
        while ( my ( $class, $count ) = each %$ref ) {
            $feature_count->{$feature}                    += $count;
            $class_count->{$class}                        += $count;
            $co_occur_count->{ $class . "\t" . $feature } += $count;