Analisis Segmentasi Demografi Pembeli Smartphone pada Kalangan Mahasiswa di Kabupaten Jember Menggunakan Metode K-Means Clustering

Abstract

This study aims to implement the K-Means method in mapping Android and Iphone Smartphone users among students of Jember University. The variables used in this study are age, gender, family size, ethnicity, major, and monthly expenditure as demographic variables, and the variables for Smartphone selection are quality, price, performance, durability, reliability, brand, features, design, and Smartphone type. The preprocessing technique used in this study is data transformation, and the method used is Label Encoder. The measurement of distance in determining clusters in this study uses the Euclidean Distance method. To measure the optimal cluster value, this study uses the Elbow method and Silhouette Coefficient. From the implementation of the method and analysis of 150 student data, based on the Elbow method measurement, the best k cluster value is 3 with SSE value of 2522, and based on the Silhouette Coefficient method measurement, the best k cluster value is 9 with a value of 0.8037.