此算法的主要作用：屏幕上很多的點(diǎn)，把相鄰的點(diǎn)聚到離他最近的點(diǎn)。

k-means algorithm算法是一個(gè)聚類算法，把n個(gè)對象根據(jù)他們的屬性分為k個(gè)分割，k < n。它與處理混合正態(tài)分布的最大期望算法很相似，因?yàn)樗麄兌荚噲D找到數(shù)據(jù)中自然聚類的中心。

php實(shí)現(xiàn)算法代碼如下：

html' target='_blank'>class Cluster {   public $points;   public $avgPoint;   function calculateAverage($maxX, $maxY)   {     if (count($this->points)==0)     {         $this->avgPoint->x = rand(0, $maxX);         $this->avgPoint->y =  rand(0,$maxY);         //we didn't get any clues at all :( lets just randomize and hope for better...         return;     }      foreach($this->points as $p)         {          $xsum += $p->x;          $ysum += $p->y;         }         $count = count($this->points);       $this->avgPoint->x =  $xsum / $count;       $this->avgPoint->y =  $ysum / $count;   } }   class Point {   public $x;   public $y;   function getDistance($p)         {          $x1 = $this->x - $p->x;          $y1 = $this->y - $p->y;          return sqrt($x1*$x1 + $y1*$y1);         } }   function distributeOverClusters($k, $arr) {  foreach($arr as $p)         { if ($p->x > $maxX)                 $maxX = $p->x;           if ($p->y > $maxY)                 $maxY = $p->y;         }   $clusters = array();   for($i = 0; $i < $k; $i++)         {          $clusters[] = new Cluster();          $tmpP = new Point();          $tmpP->x=rand(0,$maxX);          $tmpP->y=rand(0,$maxY);          $clusters[$i]->avgPoint = $tmpP;         }   #deploy points to closest center.   #recalculate centers   for ($a = 0; $a < 200; $a++) # run it 200 times   {         foreach($clusters as $cluster)                 $cluster->points = array(); //reinitialize         foreach($arr as $pnt)         {            $bestcluster=$clusters[0];            $bestdist = $clusters[0]->avgPoint->getDistance($pnt);              foreach($clusters as $cluster)                 {                         if ($cluster->avgPoint->getDistance($pnt) < $bestdist)                         {                                 $bestcluster = $cluster;                                 $bestdist = $cluster->avgPoint->getDistance($pnt);                         }                 }                 $bestcluster->points[] = $pnt;//add the point to the best cluster.         }         //recalculate the centers.         foreach($clusters as $cluster)                 $cluster->calculateAverage($maxX, $maxY);     }   return $clusters; }   $p = new Point(); $p->x = 2; $p->y = 2; $p2 = new Point(); $p2->x = 3; $p2->y = 2; $p3 = new  Point(); $p3->x = 8; $p3->y = 2; $arr[] = $p; $arr[] = $p2; $arr[] = $p3; var_dump(distributeOverClusters(2, $arr));

鄭重聲明：本文版權(quán)歸原作者所有，轉(zhuǎn)載文章僅為傳播更多信息之目的，如作者信息標(biāo)記有誤，請第一時(shí)間聯(lián)系我們修改或刪除，多謝。