ABSTRACT
Heart disease is the most dangerous diseases among the people worldwide. It can diagnose earlier using medical history, but this method has been found to be unreliable. Earlier identification of heart disease accurately and promptly will increase the preventing heart failures. Manual methods for diagnosing cardiac disease must be more accurate and subject to inter-examiner variability. Machine learning techniques are more effective and dependable in classifying heart diseases or not. Previous studies had several drawbacks, like computing slow processes, sometimes speedy but inaccurate. To overcome this, we proposed an Ensemble classifier to achieve classification accuracy. In this study, we analyzed numerous machine learning algorithms utilized for predicting and identifying the heart diseases. The proposed mechanism involves data preprocessing, and the resultant data is applied to the proposed system for predicting heart disease. The experimental work determines the enhanced performance of the proposed system in the aspect of ROC, AUC, recall, precision, F-measure, and accuracy. The proposed Ensemble classifier has achieved 98.8% accuracy in predicting heart diseases, whereas existing approaches such XGBoost [Farhat et al.2022] has achieved 90%, and Smote-XGboost [Ishaq et al. 2021] has achieved 85.6% of accuracy.
