PREDIKSI GANGGUAN TIDUR PADA USIA PRODUKTIF MENGGUNAKAN METODE FUZZY MAMDANI

Abstract

Sleep disorders are a common health issue among individuals in their productive age, yet they often go undetected due to their subjective and varied symptoms. Lifestyle factors such as sleep duration and quality, stress level, body mass index (BMI), and daily physical activity play a significant role in influencing one’s sleep condition. This study aims to develop a sleep disorder prediction system based on the Mamdani Fuzzy Inference System, which is capable of handling data uncertainty and providing interpretable linguistic outputs. The system utilizes five input variables and produces predictions classified into three sleep disorder categories: None, Sleep Apnea, and Insomnia. Evaluation was conducted by comparing the predicted results with actual labels using a confusion matrix and classification metrics, including accuracy. The results showed that the system achieved an accuracy of 71%, with most predictions falling into the None category. This study demonstrates that the Mamdani Fuzzy method can serve as an alternative approach for predicting sleep disorders based on lifestyle data, offering advantages in both system interpretability and flexibility to support early detection of sleep disorders among productive-age individuals.