README

Installation via composer

composer require medansoftware/kmeans-algorithm-php

How to use

Example 1 Video Refrence

$Kmeans = new \Algorithm\Kmeans;
$Kmeans->setAttributes(array(
	'A', 'B'
));

$Kmeans->setDataFromArgs(1, 1);  // "First Argument" assign to "First Attribute" and "Second Argument" assign to "Second Attribute" etc.
$Kmeans->setDataFromArgs(2, 1);
$Kmeans->setDataFromArgs(4, 3);
$Kmeans->setDataFromArgs(5, 4);

$Kmeans->setClusterCount(2); // Set amount of cluster
$Kmeans->setCentroid(0, 1); // Choose centroid from array key or left blank will auto fill

// Looping for iteration
$i = 1;
for (; ; ) {
	$Kmeans->setIteration($i);
	$Kmeans->run();
	if ($Kmeans->isDone()) {

		echo "<pre>";
		print_r ($Kmeans->getCentroid());
		echo "</pre>";

		echo 'Iteration ended on : '.$Kmeans->countIterations();

		echo "<pre>";
		print_r ($Kmeans->getAllResults());
		echo "</pre>";
		break;
	}
	$i++;
}

Example 2 Video Refrence

$Kmeans = new \Algorithm\Kmeans;
$Kmeans->setAttributes(array(
	'x'
));

$Kmeans->setDataFromArgs(1);
$Kmeans->setDataFromArgs(2);
$Kmeans->setDataFromArgs(6);
$Kmeans->setDataFromArgs(7);
$Kmeans->setDataFromArgs(8);
$Kmeans->setDataFromArgs(10);
$Kmeans->setDataFromArgs(15);
$Kmeans->setDataFromArgs(17);
$Kmeans->setDataFromArgs(20);

$Kmeans->setClusterCount(3); // Set amount of cluster
$Kmeans->setCentroid(1, 5, 7); // Choose centroid from array key or left blank will auto fill

// Looping for iteration
$i = 1;
for (; ; ) {
	$Kmeans->setIteration($i);
	$Kmeans->run();
	if ($Kmeans->isDone()) {

		echo "<pre>";
		print_r ($Kmeans->getCentroid());
		echo "</pre>";

		echo 'Iteration ended on : '.$Kmeans->countIterations();

		echo "<pre>";
		print_r ($Kmeans->catchLogs());
		echo "</pre>";
		break;
	}
	$i++;
}

Langkah-langkah & Catatan :

Hitung jumlah data.
Tentukan jumlah "K" atau cluster (tidak boleh lebih dari jumlah data).
Pilih titik secara acak sebanyak "K", dimana titik ini akan menjadi pusat (centroid) dari masing-masing kelompok (clusters).
Hitung jarak dan alokasikan masing-masing data ke centroid atau rata-rata terdekat.
Tentukan centroid baru / rata-rata yang ada di masing-masing cluster.
Kembali ke step 3, apabila masih ada data yang berpindah cluster atau ada perubahan nilai centroid atau jika tidak ada perubahan maka hentikan proses clustering.

Refrence :