Heart disease detection using novel ensemble approach: RF- GB-SVM stacking classifier

Cardiovascular diseases (CVDs) are among the most concerning and common health issues causing millions of deaths every year, thereby exerting extra burden on healthcare systems globally. Given that heart disease is the leading cause of death globally, there has been an urgent need to find efficient risk prediction models that would help reduce the alarmingly increasing mortality rates over the years. This paper focuses on enhancing patient results and reducing mortality through early detection and diagnosis of heart disease. Although recently proposed new models have improved performance without testing accuracy, most algorithms still suffer from overfitting issues that limit their use in a real medical facility. In this paper, we propose a novel integrated model that combines the power of three different ML algorithms, namely RF, GBM, and SVM, using logistic regression (LR)-based stacked classifiers. This combination not only achieves higher results in all performance evaluation parameters, but also improves the robustness and generalisation by reducing the overfitting of the model. The model has a precision of 96.64%, a precision of 95.56%, a recall of 98.47%, an F1 score of 96.99%, and a ROC-AUC score of 97.63%, which are better than the existing model. These results demonstrate the potential of our model to be an effective tool for early diagnosis in clinical settings, which can facilitate timely intervention and improve patient outcomes.