Pengelompokan Pola Gerakan Berdasarkan Data Akselerometer Android Menggunakan Metode K-Means untuk Penelitian Pengenalan Aktivitas: Grouping Movement Patterns Based on Android Accelerometer Data Using K-Means Method for Activity Recognition Research

Widiyanto, Muhamad
Hasbi Firmansyah

SITEDI
Vol. 2 No. 4 (2025): Jurnal Sistem Informasi dan Teknologi Digital (SITEDI)

The objective of this research is to identify human movement patterns from accelerometer data using the K-Means algorithm. The data used was collected from a mobile phone's accelerometer placed in a chest pocket, which recorded walking activity along a specific route. The analysis process began with a preprocessing stage that included data cleaning and transformation, and concluded with clustering using the K-Means algorithm. The clustering results show that the data, consisting of 5068 data points, was divided into 5 clusters. The algorithm's performance evaluation indicates that the K-Means algorithm is effective in grouping data based on movement patterns. The conclusion of this study is that the K-Means algorithm is a reliable approach for identifying movement patterns from accelerometer data and can be utilized for various applications, such as user authentication and activity monitoring.