Application of the k-means clustering method and simple linear regression to new student admissions as a promotion method

Abstract

At private label universities in Indonesia, new students are still the main thing in terms of achieving university operational income. This study intends to group the data of ITTelkom Surabaya students by utilizing the data mining process using the k-means clustering method, then the results of the clustering are forecasted using simple linear regression to be able to predict the achievement of new students as the effect variable and year as the causative variable. The results of this study consist of 5 variables, namely student province, student study program, income of student parents, student parent work and student ethnicity, each of which consists of 4 clusters, then each cluster is predicted for achievement 3 the coming year 2022,2023,2024. It can be concluded that the highest combination of student/parent student profiles was obtained from East Java province, information systems study program, parents' income of 5-10 million per month, the occupation of other parents and the ethnicity of students from Java. The highest forecasting results are found in the income variable of students' parents in cluster 3 with predictions of 1292 students in 2024. It is hoped that with clustering and forecasting based on this research, ITTelkom Surabaya can make the right decision as a basis for decision making to determine strategy in promoting the campus.