Network Modeling Analysis in Health Informatics and Bioinformatics Computer Science (miscellaneous)

Improving heart disease prediction using multi-tier ensemble model

Hera, Sabeena Yasmin, Amjad, Mohammad, Saba, Md Khalaf
Springer Nature · 2022-10-16 · DOI 10.1007/s13721-022-00381-3

Abstract

Heart disease is the leading cause of morbidity and mortality around the globe. The only way to reduce these deaths is to diagnose them at the earliest. The aim of this paper is to evaluate the effect of ensemble learning methods in creating the model for effective and improved diagnosis of Heart disease (HD). In this paper, we propose a multi-tier ensemble (MTE) model with Random Forest feature selection for effective and improved diagnosis of HD. The tiers consist of different combinations of ML and ensemble learning techniques. The model is then validated using a dataset that is curated using five widely known heart disease datasets, Cleveland and Statlog datasets to confirm the efficacy of the proposed methodology. The experimental results are evaluated in several performance measures i.e., accuracy, precision, recall and f-measure and area under curve (AUC). The results reveals that the proposed ensemble model outperformed with accuracy 93.76% as compared to the other classification models for the curated dataset. The Wilcoxon signed-rank test is also used to examine the significance of the model. The proposed model also performed significantly better than previous studies and state-of-art methods, implying that it can play an important role in medical decision-making for HD diagnosis.

Read at publisher →