Classification of the iris dataset by LogisticRegression

iris-test.php

<?php
require_once __DIR__ . '/../vendor/autoload.php';

use \Phpml\Classification\Linear\LogisticRegression;
use \Phpml\CrossValidation\StratifiedRandomSplit;
use \Phpml\Dataset\Demo\IrisDataset;
use \Phpml\Metric\Accuracy;
use \Phpml\Metric\ConfusionMatrix;

$dataset = new IrisDataset();
$dataset = new StratifiedRandomSplit($dataset, 0.2, 0);

$classifier = new LogisticRegression(
    500,
    false,
    LogisticRegression::BATCH_TRAINING,
    'log',
    'L2'
);

$classifier->train(
    $dataset->getTrainSamples(),
    $dataset->getTrainLabels()
);

$predicted = $classifier->predict(
    $dataset->getTestSamples()
);

$actual = $dataset->getTestLabels();

$accuracy = Accuracy::score($actual, $predicted);
$confmat = ConfusionMatrix::compute($actual, $predicted);

echo 'Accuracy = ', $accuracy * 100, "%\n";
foreach ($confmat as $row) {
    foreach ($row as $col) {
        printf('%3d ', $col);
    }
    echo "\n";
}