Implementation of K-Means Algorithm for Clustering Scholarship Candidates at TPA Anak Sholeh

Abstract

The Qur'an Education Park (TPA) plays an important role in improving children's education in the surrounding environment by providing facilities for learning to read and write the Quran, reading short prayers, and other positive activities. However, the criteria for scholarship recipients at TPA Anak Sholeh are still unclear, even though the provision of this scholarship aims to motivate students to be more active in learning and excelling. This study aims to determine more specific criteria for scholarship recipients, such as attendance, achievements, home conditions, and parents' income. Another objective is to validate and improve the accuracy of the K-Means Clustering model used in grouping the data of prospective scholarship recipients into three categories: accepted, considered, and rejected. The research method used is the K-Means Clustering algorithm. Algorithm performance evaluation is carried out through metrics such as accuracy, precision, recall, f-measure, and cross-validation. Testing was carried out on TPA student data as many as 80 student data. The results of the grouping using 3 clusters were obtained as many as 46 students received C1 scores, C2 scores were rejected by 12 students and C3 scores were considered for 22 students. The determination of the centroid distance has an effect on the value of the cluster results when conducting the scholarship attribute collateralization.