Predictive Modeling for Cardiovascular Disease in Patients Based on Demographic and Biometric Data

Abstract

Cardiovascular disease (CVD) remains the leading global cause of death, highlighting the urgent need for accurate risk assessment and prediction tools. Machine learning (ML) has emerged as a promising approach for CVD risk prediction, offering the potential to capture complex relationships between clinical and biometric data and patient outcomes. This study explores the application of support vector machines (SVMs), ensemble learning, and artificial neural networks (NNs) for predictive modeling of CVD in patients. The study utilizes a comprehensive dataset comprising demographic and biometric data of patients, including age, gender, blood pressure, cholesterol levels, and body mass index, features. SVMs, ensemble learning, and NNs are employed to construct predictive models based on these data. The performance of each model is evaluated using metrics such as accuracy, sensitivity, specificity, and the area under the receiver operating characteristic (ROC) curve (AUC). The results demonstrate that all three models achieve accuracy performance in predicting CVD events, with AUC values ranging from 0.85 to 0.92. Ensemble learning exhibits the highest overall accuracy, while SVM and ANN demonstrate strengths in specific aspects of prediction. The study concludes that Machine learning algorithms, particularly ensemble learning, hold significant promise for improving CVD risk assessment. The integration of ML-based predictive models into demographic practice can facilitate early intervention, personalized treatment strategies, and improved patient outcomes.