Risk Assessment of Heart Disease Based on Random Forest Model and SMCN

¹ School of Computer and Information Science School of Software, Southwest University, 400715, Chongqing, China
² School of Economics and Management, Shanghai Institute of Technology, 200235, Shanghai, China
³ School of Computer and Artificial intelligence, Hubei University of Science and Technology, 437100, XianNing, China

^* Corresponding author’s email: This email address is being protected from spambots. You need JavaScript enabled to view it.

Abstract

This study aims to compare the performance of the Random Forest model and the SMCN model in the task of heart disease risk discrimination, providing a basis for the selection of clinical intelligent screening models. The EarlyMed dataset with 70,000 samples was adopted, and 18 clinical features were selected. After data cleaning, stratified sampling and Z-score standardization, the Random Forest model and the SMCN model were constructed respectively, and Accuracy, Precision, Recall, F1-score and AUC were used as performance evaluation indicators. Both models demonstrated outstanding discriminative performance on the test set. The AUC of Random Forest and SMCN both exceeded 0.999, and the accuracy rates both exceeded 0.99. However, SMCN showed slightly better balance in the recognition of true negatives and true positivity. Both the Random Forest model and the SMCN model can be effectively used for heart disease risk discrimination. Among them, SMCN has a slight advantage in classification balance due to its heterogeneous integration strategy. This study confirmed the potential of ensemble learning models in capturing nonlinear relationships in complex medical data, which has reference value for optimizing the early screening system of heart disease.