SISTEM PAKAR BERBASIS WEB UNTUK DETEKSI DINI GANGGUAN NEUROLOGIS MENGGUNAKAN LOGIKA FUZZY MAMDANI

This study develops a web-based expert system for the early detection of neurological disorders by applying the Fuzzy Mamdani method as the core inference mechanism and Simple Additive Weighting (SAW) to determine the ranking of possible diseases based on symptom weights. The system focuses on five types of neurological disorders—Low Back Pain, Vertigo, Ischemic Stroke, Epilepsy, and Peripheral Neuropathy—which were determined through literature studies and validated by a neurology expert. The diagnostic process involves fuzzification of symptoms, evaluation of fuzzy rules, aggregation, and defuzzification using the centroid method to produce crisp values, followed by normalization and weighting with SAW to obtain the final score and ranking of the most likely diseases experienced by users. The evaluation results show that the system performs well, with a Mean Absolute Error (MAE) of 2.8%, a Root Mean Squared Error (RMSE) of 2.83%, and an F1 Score of 0.75, demonstrating consistency with manual calculations, low error rates, and high accuracy. The system is also equipped with a user-friendly interface that can be easily accessed by the general public without additional installation, making it a promising tool for supporting early detection of neurological disorders quickly and accurately, as well as serving as a decision support instrument in the medical domain.
Keywords — Expert System, Fuzzy Mamdani, Simple Additive Weighting, Neurological Diagnosis, Web.