DETERMINATION OF POTENTIAL BUSINESS LOCATIONS USING DATA MINING CLUSTERING

Potential locations for businesses are highly sought after by business people to set up, expand their business, or establish a new business. Limited information on potential business locations is still a problem faced by many business people in making business decisions. The purpose of this research is to overcome the limitations of potential business location information. The approach used is the K-Means data mining clustering method which is compared to the Gaussian Mixture Model. The dataset used is residential, road access data and business points that already exist around the location. Both clustering methods are compared to the model evaluation method to determine the model with the best performance. The results show that the clustering method with the K-Means algorithm is the clustering model with the best performance. The results of the clustering resulted in 2 clusters, one of which is a cluster of potential business locations of 1041 locations. The conclusion of this study is that data mining clustering can be used to determine the optimal business location cluster. The results of this study can be recommended for business people to look for potential business locations, and for local governments to publicize potential business locations in order to attract investors from outside.