Integrating Regression to fuzzy Based Rule Extraction for CHD Risk Prediction
Ram Kishor, Chandrashekhar Diwakar · International Journal of Technological Advancements and Industrial Applications · 2025
Coronary Heart Disease (CHD) remains a leading cause of mortality worldwide, necessitating the development of accurate risk prediction models. This study integrates regression-based rule extraction with fuzzy logic inference to enhance CHD risk assessment. Traditional fuzzy models rely on expert-defined rules, which may introduce subjectivity and inconsistency. To overcome this limitation, we employ data-driven rule generation using subtractive clustering and refine rule coefficients through least squares regression (LSR). Fractional-order fuzzy membership functions are applied to better capture the uncertainty in medical parameters, including age, LDL, cholesterol, HDL, triglycerides, and systolic pressure. The hybrid fuzzy-regression model improves predictive accuracy by optimizing rule weights based on real-world patient data. A series of 2D visualizations illustrate key relationships between risk factors and CHD likelihood. The results indicate that integrating regression significantly enhances the robustness of fuzzy rule-based models, making them more adaptive to real-world medical datasets. This approach provides a scalable, data-driven decision support system for early CHD diagnosis, potentially aiding healthcare professionals in personalized patient risk assessment.