Deteksi Dini Hipertensi Pada Perusahaan Jasa Keamanan Menggunakan Metode Decision Tree

Abstract

Hypertension poses a serious threat to employees in the security services sector, especially those requiring physical and mental readiness. Early detection of hypertension is crucial to prevent potentially fatal complications. This research aims to develop a hypertension prediction model using machine learning algorithms to enhance accuracy compared to conventional methods. Data from Medical Check-ups (MCUs) of 800 employees from PT Demitra Karsa Perdana Site Adaro Tanjung Kalimantan Selatan in 2023 were utilized. The data encompassed health factors and lifestyle choices influencing hypertension, such as age, stress history, cholesterol levels, weight gain, family medical history, smoking habits, alcohol consumption, and diagnoses. The Decision Tree algorithm was employed to classify the data. Classification results were evaluated by calculating accuracy, precision, and recall for hypertension classes (high or normal). Findings indicated that Decision Tree was the best algorithm for predicting hypertension, achieving an accuracy of 92.01%, precision of 99.41%, and recall of 84.42% for high hypertension class. However, the recall value for high hypertension class still needs improvement to prevent high-risk employees from going undetected. This research benefits PT Demitra Karsa Perdana in predicting hypertension among their employees and contributes to the advancement of science and technology. Thus, this study not only enhances our understanding of hypertension but also provides insights into the use of machine learning for early detection of serious diseases in the workplace.