Simulasi Kinerja Karyawan di Kantor Pertanahan Labuhanbatu Menggunakan Algoritma C4.5

Abstract

Employee performance analysis using the C4.5 algorithm in data mining aims to identify and classify employees based on their performance. The analysis process includes several stages, namely data analysis, preprocessing, model design in data mining, and method evaluation. From 47 sample data analyzed, the results show that 40 employees have good characters, while 7 employees have bad characters. Good employee characters are characterized by punctuality and high discipline in carrying out their duties. Conversely, bad employee characters are characterized by unpunctuality and low discipline, which have a negative impact on productivity and efficiency in the workplace. The results of this classification help identify areas that require more attention and intervention to improve overall employee performance. Model evaluation is carried out using two widgets, namely Test and Score and Confusion Matrix. The evaluation results of these two widgets show perfect accuracy of 100%. Meanwhile, the Confusion Matrix widget shows that all predictions are in accordance with the actual data without any errors in classification. These results confirm that the C4.5 algorithm is very effective and accurate in classifying employee performance. The perfection of the evaluation results shows that the C4.5 algorithm is very suitable for use as a classification model in employee performance analysis. The 100% accuracy of both widgets indicates that this algorithm is not only able to predict correctly but also consistently in various evaluation tools.