Abstract
Cardiovascular disease (CVD) is the leading cause of death globally, demanding accurate risk prediction models for early intervention and prevention. This project aimed to develop a Machine Learning (ML) model for predicting the 10-year risk of CVD. A comprehensive review of existing literature was conducted, discussing the methods, algorithms, and data sources used in different studies, to evaluate the performance of various models. The review highlighted the potential of ML for improving CVD risk assessment, and the challenges and limitations of current research.
The UCI Heart dataset served as the training data for various ML models, including Logistic Regression (LR), Decision Trees (DT), Random Forests (RF), Support Vector Machines (SVM), Artificial Neural Networks (ANN), and K-Nearest Neighbors (KNN). To optimize model performance, Cross Validation (CV), normalization techniques, and hyperparameter tuning were employed. We report the results, comparing them with traditional models.
The implications of this research extend to improved preventive strategies and interventions, potentially alleviating the burden of CVD on individuals and healthcare systems by more targeted interventions, and the optimization of healthcare resources.