An Algorithm for Discretization of Real Value Attributes Based on Interval Similarity

Algorithm 2

Step1: Compute inconsistency rate of information system;

Step2: Sort data in ascending order for each attribute and calculate the similar

    value SIM of each adjacent intervals according to (5) and (6);

Step3: Merge

While (merge-able cut point)

{

 Search cut point that has the maximal similar value, then merging it;

 If (many maximum values)

   {

  Merge adjacent two intervals with the smallest number of classes;

  If (Incon_rate increases)

  {

   Withdraw merging;

   Exit procedure;

     }

  Else {break; goto Step2;}

   }

 If (some several maximum values and the same class number of classes

    among groups of adjacent intervals)

     {

   Merge the adjacent two intervals with the smallest number of samples

   of adjacent intervals;

   If (Incon_rate increases)

       {

   withdraw merging;

   exit procedure;

   }

       Else {break; goto step2;}

  }

}