Performance Analysis of Subtractive Clustering Algorithm in Determining the Number and Position of Cluster Centers

The basic concept of the subtractive clustering algorithm is to choose a data point that has highest density (potential) in space (variable) as center cluster. number and position cluster centers formed are influenced by given radius (r) parameter value. If value very small, it will result neglect potential points around too large, increases contribution all points, thereby canceling effect density. determined based on iteration process finding with neighbors. This study uses partition determine (candidate center) be selected generating clustering. From experiments have been carried out 4 datasets, results obtained, for dataset 1 average fuzzy silhouette 0.35 0.9088 clusters 2. While 2, 0.40 0.6742 3. 3, 0.50 0.7434 last fourth dataset, 0.6630 subscractive widely applied fields transportation, GIS, big data, control electric voltages, electrical energy needs, knowing area population health such breast cancer diagnosis, which related needs human life.