An Ensemble Cognitive Model for Stroke Prediction Using Unstructured Health Information Powered by Machine Learning

Abstract

Machine Learning (ML) algorithms have procured a profound position in healthcare sectors, especially in diagnosis, treatments, and recommendation systems. The ML is evolving as an aiding tool for medical practitioners in disease diagnosis. Also, the feature selection reveals the latent relationships among the features, which emerge significant scope for clinical research. In the proposed study, a cognitive ensemble model (CEM) was developed to predict the probability of stroke among various subjects using highly raw clinical data. The optimal base learners are made in such a way that each of them complements one another. The proposed CEM is tested on a real-world dataset on important classification metrics. The results indicate that the CEM deployed in the healthcare sector forewarns patients regarding the probability of stroke.