Segmentation model toward promotion target determination using k-means algorithm and Elbow method

Promotion is one way that can be used to increase the number of registrants at a university. The number of registrants for a university has greatly increased in the student segmentation process carried out to determine the target area for promotion. A good segmentation process can produce accurate information that can be used as a basic promotion or marketing strategy. This study develops a model based on the KMeans algorithm for the segmentation process based on the number of registrants and the citizen population (school-age) in a city. This algorithm was chosen because it can group data that has similarities into a cluster. Elbow method is used to determine the cluster optimum number based on Within Cluster Sum of Squared Errors (WSS) calculations and inertia is used for calculations to visualize data in a plot or graph. The results of this study are in the form of a model that can interpret clusters, namely clusters with high population but have a small number of registrants, cluster with high population and have a high number of registrants (promotional targets), cluster with a low population but have high number of registrants, and a cluster with low population but less registrants. With the interpretation of the model a university can determine the target city for promotion based on the targeted cluster.