In fault feature selection, if each location contains the least features and the highest accuracy, it is considered as the best. Each location is evaluated by a developed accuracy function that uses the KNN classifier calculator for classification accuracy and the number of selected features