ENSEMBLE LEARNING FOR HEART DISEASE PREDICTION: A NOVEL APPROACH USING ACCURACY-BASED WEIGHTED AGING CLASSIFIER ENSEMBLE
Abstract
The early detection of the risk for heart disease is crucial for the prevention of cardiovascular diseases which is still the leading cause of death globally. This paper introduces an ICIT2F mean partitioning-based CART model for heart disease classification if integrated with four others models. In this method, the entire dataset is divided into smaller chunks, and a CART classifier is trained on each chunk. Then, an Accuracy Based Weighted Aging Classifiers Ensemble (AB-WAE) is used to combine the classifiers, individually, creating a homogeneous ensemble with high prediction power. Proposed approach is tested on the Cleveland and Framingham heart disease databases with 93% and 91% classification accuracy respectively. A comparison with other state-of-art popular machine learning algorithms such as random forest, gradient boosting, and support vector machine shows that the proposed method outperforms existing classical methods. Further ROC curve analysis confirms its robust discriminative ability. In conclusion, the present work provides a solid and accurate tool for heart disease risk prediction which could assist clinicians in early diagnosis, timely intervention and better patient management.