Diabetes Level Based Heart Disease Risk Prediction System

One of the major diseases that cause ill health worldwide is cardiovascular disease, particularly for people who already have diabetes, as their risk of developing heart disease is considerably higher. The risk increases further with factors such as high blood pressure, high cholesterol, obesity, or a sedentary lifestyle. Identifying the risk at an early stage gives patients and doctors an opportunity to take necessary action before serious complications arise. This paper presents a desktop application using machine learning to predict the risk of heart disease in people with diabetes. The system uses health indicators such as age, sex, blood sugar, HbA1c, blood pressure, cholesterol, body mass index, heart rate, lifestyle indicators, and symptoms. The dataset is preprocess by removing duplicate rows, filling missing values, handling extreme values, and standardizing numeric columns. Six classifier models are applied to compare their performance. The system provides the risk percentage and an easy-to-understand risk classification. It also supports patient data management, prediction history, graphical analysis, model comparison, and PDF report generation. Patients can access and update their records when required. The system provides a complete workflow from data gathering and analysis to decision making and documentation. It is designed as a learning tool and is not a medical device or a substitute for professional medical examination.