Description of Invention
Diabetes Mellitus is a significant global health concern, contributing to chronic diseases such as heart disease. Individuals with diabetes are approximately 2 to 3 times more likely to develop cardiovascular conditions, with nearly one in three diabetic patients affected. Traditional diagnostic procedures often lack timely and accurate risk prediction, necessitating machine learning-based solutions. This study applies machine learning models to predict heart disease risk in diabetes patients. By integrating data-driven feature selection with advanced model tuning, the approach enhances prediction accuracy. Among the seven algorithms used, CatBoost and LightGBM achieved the strongest performance, demonstrating potential for early diagnosis and personalized healthcare.