• AI::NNFlex::Dataset::new
• AI::NNFlex::Datasets::run
• AI::NNFlex::Dataset::learn

AI::NNFlex::Dataset::new

Constructor, creates a data set to run through the network. Takes as parameters an anonymous hash containing input array and target output array, as follows:

my $dataset = AI::NNFlex::Dataset->new([
	[0,1,1,0,1,0],[1,1],
	[0,0,1,1,0,0],[1,0]]);

returns dataset object

AI::NNFlex::Datasets::run

Run a data set through the network and return the outputs.

Takes as parameter the network object you want to run the data against.

Inspired by the Xerion/UTS approach to managing datasets.

Syntax:

$dataset->run($network);

returns a reference to an hash containing input to actual outputs in array form.

AI::NNFlex::Dataset::learn

for each entry in the dataset, run through the network then call the networks learning algorithm. Takes as a parameter the network you want to train.

Syntax:

my $error = $dataset->learn($network);

returns network sqrd err.