CARDIAC RISK ASSESSMENT FROM CLINICAL DATA

Cardiovascular diseases (CVDs) are one of the leading causes of death worldwide, causing millions
of deaths each year. These include coronary artery disease, heart attacks, and strokes, mainly due to
unhealthy lifestyle habits and genetic factors. Early detection is essential to reduce mortality and improve
patient outcomes. However, traditional diagnostic methods are expensive, time-consuming, and less
accessible in rural areas. Machine Learning (ML) has emerged as an effective solution for early disease
prediction in healthcare. It can analyze large amounts of medical data and identify hidden patterns. This
project proposes a Machine Learningbased Cardiac Risk Assessment System to predict heart disease using
clinical data. The system uses supervised algorithms such as Logistic Regression, Decision Tree, Random
Forest, KNN, and SVM. The models are trained using the Cleveland Heart Disease dataset with 303 records
and 13 features. Data preprocessing techniques like normalization, encoding, and handling missing values
are applied. The models are evaluated using metrics such as accuracy, precision, recall, F1score, and ROC
AUC. Among them, Random Forest provides the best performance. The system is implemented using Flask
as a web application for real-time prediction.