A Mamdani Fuzzy Inference System for Comprehensive Diabetes Risk Evaluation

Diabetes mellitus is a chronic condition characterized by insufficient insulin production and the inability to effectively utilize the insulin hormone, leading to elevated blood glucose levels. As diabetes is incurable and persists throughout a patient's lifetime, effective management requires early identification of risk factors to ensure continuous and feasible treatment. This study proposes a unique decision support system based on the Mamdani Fuzzy Inference System (MFIS) to predict diabetes risk. This study utilizes a publicly available diabetes dataset containing 442 data that includes nine key health metrics such as BMI, blood glucose, cholesterol levels, and blood pressure to predict the risk of diabetes. The dataset aims to support early diagnosis and better diabetes management by providing insights into individual risk factors. This resource contributes to the development of more accurate decision-support systems in healthcare. We identify key variables and establish a membership function to assess individual risk. The Fuzzy Inference System (FIS) is well-suited for handling complex and uncertain conditions, making it an effective tool for this application. The advantages of the MFIS include its ability to incorporate expert knowledge through linguistic variables and fuzzy rules. This system demonstrates the practical benefits of MFIS in developing intelligent decision-making tools, particularly in the domain of diabetes risk prediction.