The transmigration program implemented by the government is not only for the equal distribution of population, but also for economic development and inter-regional development in Indonesia. Malang Regency is currently still opening a transmigration program, but many potential areas for transmigration in Malang Regency are not comparable with the transmigration quota provided by the government as well as the community's interest in transmitting transmigration. Therefore, it is necessary to group regions in Malang Regency that have potential to transmigrate in order the government can provide accurate information for people in Malang Regency and could see the potential of each region. In this study the method used is Improved k-Means which is able to improve the accuracy of clustering results and the quality of clustering when compared with the k-Means method. It is because the initial centroid selection in the Improved k-Means method is done by selecting value of the farthest distance between data. Based on the results of testing conducted the Improved k-Means method produces better cluster quality compared to the k-Means method, in the silhouette coefficient value of 0.29135900558873151 with parameters of the number clusters of 3 and maximum iteration of 3. Clusters 1 there are 4 districts with no potential transmigration, cluster 2 there are 27 districts that have the potential to transmigrate, and cluster 3 there are 2 districts which have the potential to transmigrate.
Copyrights © 2020