Clustering Analysis of Admission of New Students Using K-Means Clustering and K-Medoids Algorithms to Increase Campus Marketing Potential

Abstract

Acceptance of new students is a very important activity for a high school or university. The admissions data has not been utilized by the campus in making strategic decisions, marketing potential, and considering invitations through academic admissions. So, to assist in processing the new student admissions data, in this study the design and analysis of new student admissions data was carried out using stages in data mining. The clustering method approach can be applied in analyzing the potential level of PMB quality produced by utilizing the PMB recording dataset for the 2023 period. 86 data records. The K-Means and K-Medoids algorithm models that are applied have results that show a new insight, namely grouping based on 2 clusters, cluster 1 (C0) is a pass category while cluster 2 (C1) has not been determined. The results of the K-Medoids algorithm which has cluster 1 (C0) 60 results, cluster 2 (C1) has 26 results is a potential pass of 60 and has not yet been determined 26 of the data tested 86 while the results of the K-Means cluster 1 algorithm (C0) 40 , cluster 2 ( C1 ) 46 is a potential pass consisting of 40 and 46 undetermined data from the 86 datasets tested. Testing using the RapidMiner Studio application can also produce similar insights, namely each cluster has Davies Bouldin Index or DBI results from each K-Means and K-Medoids algorithm. K-Means has a Davies Bouldin Index result of -0.533 while K-Medoids has a Davies Bouldin Index result of -0.877