Analisa Model Clustering Untuk Pemetaan Kualitas Lulusan Mahasiswa Berdasarkan Dataset Tracer Study

Abstract

Graduate data from the tracer study process is critical in assessing the quality of a university's graduates. From this data, universities can see an objective picture to measure and evaluate the curriculum, materials, and the achievement of learning competencies so far whether they are following what is expected by graduate users. This will provide input to university management in making strategies and policies to improve quality. However, the problem is that the amount of data available so far has not been maximized properly to assist management in making decisions. Data on graduates and users of existing graduates are only processed into semester and annual reports and there is no in-depth analysis. So management does not get information that helps improve graduates' quality in the future. Optimization of clustering methods using the elbow method with a comparison of other distance formulas such as Euclidean Distance, Mahalanobis Distance, and Manhattan City Distance to improve the performance of mapping results. The DBI result obtained is 1.89 for the number of 6 clusters.