PENERAPAN ALGORITMA K-MEANS CLUSTERING PADA K-HARMONIC MEANS UNTUK SCHEDULE PREVENTIVE MAINTENANCE SERVICE

Sari

Abstract: Vehicle  maintenance  is a  very  important  sector  in  terms  of  economy  and  safety,  a good understanding of vehicle maintenance is very important  from the owner of the vehicle it self or from the company. Maintenance for the vehicle is considered as part structure of the activity in a series of improvements, as well as a planned activity to prevent potential errors resulting in damage. Schedule preventive maintenance is one of the methods that are used for vehicle maintenance scheduling. SPM is widely used because  it  can  determine  component  reliability  item,  so  as  to  reduce  the  cost  of repairs,  but  this  method  has  the  disadvantage  that  reparations  are  made  to  the  unit item could potentially breakdown, as well as the application of SPM only on certain types  of  vehicles.  To  solve  this  problem  it  is  proposed  the  one  application  of  a method,  algorithm  K-Means  Clustering  is  one  of  the  methods  to  be  applied  in  the schedule vehicle maintenance services, K-Means algorithm is widely used because it is  easy  and  simple.  From  the  models  created  will  then  be  tested  using  Confucion Matrix to determine how the level of accuracy, and describes the results of a positive predictive accuracy results are correct, the positive predictions were wrong, negative predictions  are  true,  and  false  negative  predictions.  From  these  experiments  showed that  the  application  of  K-Means  Clustering  algorithms  in  the  vehicle's  maintenance schedule capable of generating predictive value and accuracy that is optimal by 70%.