Application of K-Means algorithm with CRISP-DM method in student data analysis as a support for promotion strategy

Abstract

Every year the number of prospective students in a high school is uncertain, this can be seen based on historical data taken within the last five years, namely from 2015-2020. In 2016 there was an increase in the number of students, but in 2017-2018 it decreased, then the number of students increased in 2019 and in 2020 decreased. This is due to several factors, one of which is the type of promotion carried out by high schools. In this case, it is certainly one of the tasks that must be done by the high school how to determine the right promotion strategy to increase the number of prospective students. This study aims to group student data using the k-means algorithm as an effort to determine campus promotion strategies. By using the k-means algorithm, student data can be grouped based on specified criteria. The method that has been applied in this study is the K-means algorithm, because the k-means algorithm is a very popular method used by previous researchers to explore information used to determine promotional strategies. Data collection was carried out by literature studies, interviews, documentation and the CRISP- DM method. At the data processing stage using the RapidMiner application. The results obtained from research on the application of the k-means algorithm to student data analysis were obtained one cluster that could be used as a reference for determining campus promotion strategies and three types of promotional strategies were obtained, namely school visits, brochures, and the internet.