pawa- / NaiveBayes-Classifier

多項モデル＋最大事後確率推定のナイーブベイズ分類器

NAME
    Algorithm::MyNaiveBayes - Oreore NaiveBayes Classifier

SYNOPSIS
      use Algorithm::MyNaiveBayes;
      my $nb = Algorithm::MyNaiveBayes->new(
          instances_path  => '/tmp/mnb_instances'.
          classifier_path => '/tmp/mnb_classifier',
      );

      #$nb->init; # delete previous data

      $nb->add_instance(
          label      => 'plus',
          attributes => { good => 3, bad => 1 },
      );

      $nb->add_instance(
          label      => 'minus',
          attributes => { bad => 1, boring => 3 },
      );

      # ...

      $nb->train;

      my $result = $nb->classify(
          attributes => { good => 2, bad => 1, boring => 1 }
      );

DESCRIPTION
    Algorithm::MyNaiveBayes is Oreore NaiveBayes Classifier. This uses
    multinominal model. In smoothing, maximum a posteriori estimation is
    used. I think Algorithm::NaiveBayes on CPAN is better than this module.

AUTHOR
    pawa- <pawa[at]dojikko.com>

SEE ALSO
    Algorithm::NaiveBayes

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