Index Based Hidden Outlier Detection in Metric Space

Algorithm 5

Index based hidden outlier detection algorithm.

Input: , ,

Output:

(1)  ; ; ; // is a data block of dataset

(2)  pivot pivotSelection();

(3)  (); // is a data block of dataset

(4)  while  

(5)   if (Theorem 10 holds for ) then break;

(6)   else

(7)    ;

(8)    ;

(9)    (L.id, startID);

(10)    for each in with order

(11)    for each in   

(12)      if (Theorem 11 doesn’t hold for and )  

(13)       ; update .NN

(14)       if   remove from ;

(15)    for each in   

(16)    if (Theorem 9 doesn’t holds for )

(17)     ()  //update ;

(18)      //update ;

(19)    for each cs in   

(20)     if ()

(21)      delete cs from ;

(22)    if ()

(23)     insert to ;

(24)   process();

(25)  return ;