Cardiovascular disease prediction: a machine learning-based approach

Abstract

Cardiovascular disease is a common condition that frequently results in deadly consequences, affecting a significant number of people in their middle or later years. Globally, cardiovascular diseases are considered the deadliest illnesses, contributing to the highest death rates. Heart palpitations, nausea, and chest pain are common signs of cardiovascular disease. Significant risk factors for cardiovascular disease include age, gender, high blood pressure, stress, improper lifestyle choices, and family history. This study used eight machine learning classifiers to make predictions about cardiovascular disease more accurate. These were support vector machines, random forests, decision trees, gradient boosting, K-nearest neighbors, Gaussian Naive Bayes, MLP, and logistic regression. In the heart disorders dataset, the K-Nearest Neighbors model performed the best, achieving 86% accuracy, 86% precision, 90% recall, an 87% F1-score, and a ROC AUC value of 0.8909 in the cardiovascular disease dataset. In the healthcare sector, machine learning (ML)-based prediction models offer a more efficient way to support patient diagnosis.