Application analysis of machine learning and data visualization in heart failure prediction

Abstract

Heart failure is a non-eligible global health challenge, characterized by increased morbidity, mortality, and health expenditures. Early detection of heart failure can help prevent disease progression and improve outcomes. Logistic regression is a machine learning technique widely used in binary classification problems. In this paper, the patient record dataset was used to predict heart failure using logistic regression. The most important features for predicting heart failure are also determined through detailed analysis, mainly including age, serum creatinine, and ejection fraction. The results suggest that logistic regression can be a valuable tool for predicting heart failure and improving patient outcomes. Further research could explore other machine learning algorithms and more sophisticated feature selection techniques to further improve the prediction of heart failure.