Cardiovascular Disease Prediction Using Extreme Gradient Boosting Algorithm

Cardiovascular disease prediction using machine learning involves algorithms that can accurately predict an individual’s risk of heart disease based on their medical history and other risk factors. The current paper introduces a basic effective system designed to diagnose heart disease, which utilizes the Extreme Gradient Boosting technique using a feature selection technique. The system can predict heart disease accurately with a large amount of data. To ensure the effectiveness of the classifier, proper hyperparameter tuning is critical. In this regard, we employed random search, an ideal approach for hyper-parameter optimization, to improve the hyper-parameters of XGBoost. Moreover, the model testing technique called cross-validation was employed to check the efficiency of the prediction of the forecast model. To analyze the proposed algorithm, the researchers compared it to other algorithms. The proposed model has an area under the ROC score of 0.98 and an accuracy of 98 %, outperforming KNN, Naïve Bayes, SVM, Random Forest algorithm in terms of accuracy, positive and negative rates of the model according to the experimental results.