Application of Machine Learning Model for Predicting Mortality Rate in Heart Failure

Abstract

Heart failure is among the prevalent illnesses that can give rise to perilous circumstances. Approximately 26 million individuals experience this condition annually. According to viewpoints of cardiac experts and surgeons, the timely prediction of heart failure remains a highly intricate endeavor. Fortunately, the emergence of categorization and anticipatory models provides a promising avenue for aiding the medical sector by showcasing efficient utilization of medical information. Within this study, the aim is to refine the accuracy of prognosticating heart failure risk by employing datasets derived from clinical records. To achieve this, this paper employs various machine learning techniques to interpret the information and forecast the likelihood of elevated occurrences in healthcare repositories. This study utilizes three distinct machine learning models: decision tree, logistic regression, and random forest to derive predictive accuracies for each model. The outcomes disclose that the decision tree attains the highest predictive accuracy, reaching 88.8%. Moreover, the results and comparative analyses indicate that the present study enhances the precision of earlier heart disease predictions. Incorporating the models presented in this study into healthcare systems has the potential to enable real-time forecasting of heart failure or other medical conditions using data gathered directly from patients.