Implementasi Data Mining Metode K-Means Menggunakan Framework Python Dalam Mengelompokkan Pegawai Berdasarkan Data Presensi

Abstract

Computerized employee attendance data collection provides convenience in terms of real-time monitoring. The computerized attendance data collection process has been widely implemented by government and private agencies. With a computerized system, the data collection process is easier to carry out. The data that has been collected increases as time passes. To improve discipline in employee attendance patterns. Perumdam Tirta Bengkayang implements an attendance application with features consisting of working hours, shifts, attendance location and is equipped with documentation of incoming and outgoing absences in the form of selfies attached to the application. The data that has been collected is then analyzed to assess employee discipline levels using data mining techniques, namely the K-Means method. Data mining methods are used to group employee attendance data patterns. Data mining is the process of collecting data, and finding patterns or relationships between data. The K-Means method works by dividing data into k closest clusters. The calculation begins by determining the value of k, centroid, and closest point. Meanwhile, the analysis uses the Python library by importing the necessary libraries such as numpy, pandas, matplotlib, sklearn. Based on the results of the analysis and grouping of employees, 26.76% of employees fall into cluster 0, namely the low level of discipline, 71.83% of employees fall into cluster 1, namely the medium level of discipline, and 1.41% of employees fall into cluster 2, namely the high level of discipline.