Cardiovascular Disease Prediction Using Machine Learning

Abstract

The leading cause of death worldwide is heart disease. An effective hybrid classifier model is finally constructed to classify records and produce predictions or identifications based on significant input factors. The findings of this study lower healthcare costs and enable cardiologists to diagnose heart disease more reliably. In this context, an adaptive voting classifier is a type of ensemble learning method that combines the predictions of multiple classifiers to improve accuracy and robustness. This paper presents a heart disease prediction model based on a voting classifier, which combines the predictions of individual classifiers: a decision tree, support vector machine (SVM), k-nearest neighbors (KNN) classifier, Random Forest, and XGBoost. The models used in this study will also be helpful in situations when many patients show up daily. The application would use a few attributes about the patient’s physical state and medical history. On evaluating the proposed adaptive voting-based feature selection for classification has attained an accuracy of 99.83%, and the model has outperformed compared to the other existing state-of-art models considered in the evaluation.