Analysis of Cardiovascular Disease Prediction Using Various Machine Learning and Deep Learning Algorithms

According to the World Health Organization, cardiovascular disease (CVD) continues to claim a shocking number of lives over the world. To prevent fatalities, an accurate model with the highest level of precision is required. The emerging field of information, communication, and technology includes the Internet of Things, machine learning, and deep learning. These techniques can be used in medically assisted environments to save the lives of millions of people. The goal of this study is to analyze available data on cardiovascular diseases and predict heart disease at an earlier stage and prevent it from occurring. The research proposes several heart disease prediction models using machine learning, ensemble learning, hybridization, optimization, remote healthcare monitoring and neutrosophic techniques. The proposed regression technique model produced the maximum R-Squared value by SVM with 99.23 and RMSE by DT with 5.234. The classification technique for BO-SVM model outperformed with 93.30%, 100%, and 80% for accuracy, precision, and sensitivity, respectively, whereas the PSOGD optimization algorithm achieved the highest accuracy of 99.02%, a sensitivity of 99.01%, and a specificity of 99.15% for the Framingham dataset and an accuracy of 99.92%, sensitivity of 99.00%, and specificity of 99.26% for the Cleveland dataset. Similarly, the Meta HGBoost algorithm achieved an accuracy of 96.15% without cross-validation and 94.76% with cross-validation for Ensemble technique. The findings demonstrate that the suggested models can accurately predict the risk of heart disease.