Disease Risk Level Prediction Using Ensemble Classifiers: An Algorithmic Analysis

Abstract

Disease risk level prediction schemes have turned out to be a significant part of medical decision-making. They offer a simple and quick mode to review a patient’s risk in a specified disease that can then assist the diagnosis process. Disease prediction plays an important role in clinical applications and decision making, and also it is necessary to diagnose the disease at an early stage. In this research, a new risk level prediction model is introduced that involves 2 major phases: feature extraction, prediction. Higher-order statistical features along with approximate entropy features and singular value decomposition (SVD) entropy features are extracted. As the next process, prediction takes place by Ensemble Classifier that includes Nueral Network, Fuzzy logic, and optimized Deep Belief Network. Here, the weights of DBN are optimized via Rain Leveraged Dynamic Butterfly Optimization (RLDBO). Finally, analysis is carried out with respect to varied classification models.