Improving Accurate Prediction in Cardiovascular Disease and Optimizing Using Novel Nesterov Accelerated Gradient in Comparison with Gradient Boosting Classification

Abstract. Traditional approaches and machine learning (ML) incorporate the neural network process in predicting all medical complications and detecting the severity of health issues, mostly considered critical conditions. Heart diseases are common nowadays in all generations and occur in clinical challenges in all countries. Cardiovascular is one kind of heart disease that specialists and experts handle to diagnose in the early stages. Several ML classifiers were implemented and some produced results as incorrect diagnoses addressed as inaccurate in the medical field. The proposed work novelty Nesterov Accelerated Gradient approach produces optimal prediction due to finding the severity of cardiovascular disease. Data are preprocessed using the classification of severity in heart diseases using the Novel Gradient Boosting Classification (NGBC) algorithm which has the same classes observed manually and systematically. From the classification, output was achieved to categorize the classes that can obtain an accuracy of 94.8% accuracy based on training the model and testing in prediction. To improve performance analysis and suggest diagnosis, the proposed algorithm produces precision, recall, and confusion matrix metrics of heart disease using Novel Nesterov Accelerated Gradient (NNAG) compared with XGboost and Aquila optimizer. The better optimization demonstrates how healthcare based on heart diseases can be effectively handled at an earlier stage to save human lives.