A Hybrid Approach for Heart Disease Prediction using Genetic Algorithm and SVM

Heart disease (HD) is a critical condition that indicates dysfunctionality of the heart. It is alarming to note that the projected number of deaths caused by cardiac disorders is estimated to reach a staggering 24.2 million by the year 2030. It is crucial to have precise predictions and diagnoses for early prevention, detection, and treatment of cardiac disorders. The medical devices like electrocardiograms and computed tomography scan which are being used for identifying HD can be a major challenge due to their high cost and feasibility issues. This paper presents a machine learning-based technique for early prediction of Cardiac dysfunctionality to reduce the death rate. The proposed technique is dedicated to accurately predicting HD by leveraging the power of genetic and support vector machine (SVM) models. Through the integration of a genetic algorithm (GA), the research successfully optimized parameters and feature subset selection for SVM to provide invaluable insights for the early detection of HD. The efficiency of the proposed model has been thoroughly evaluated through rigorous experiments conducted on the renowned University of California Irvine HD dataset. The experimental results indicate that the proposed GA-SVM model achieved an accuracy of 98.0%. Furthermore, it outperformed other comparable works in terms of accuracy.