Neutrosophic Hybrid Machine Learning Algorithm for Diabetes Disease Prediction

A. Bermúdez del Sol^1,*, Edison Sotalin Nivela¹, Edwin Miranda Solis¹, Yasser H. Elawady²

¹Universidad Regional Autónoma de los Andes, Ecuador

²Computer Engineering Department, Misr Higher Institute for Engineering and Technology, Mansoura, Egypt

Emails: ua.abdelbermudez@uniandes.edu.ec; us.medicina@uniandes.edu.ec; ua.edwinmiranda@uniandes.edu.ec; y.alawady@engmet.edu.eg

Abstract

Because of its far-reaching effects, diabetes remains a major health problem on a worldwide scale. It's a metabolic illness that causes hyperglycemia and a host of other health issues, including cardiovascular disease, renal failure, and neuropathy. Many scientists have spent time and energy over the years trying to develop a reliable diabetes prediction model. Researchers are forced to adopt big data analytics and machine learning (ML)-based methodologies since there are still major open research concerns in this area owing to a lack of acceptable data sets and prediction techniques. This study seeks solutions by way of an examination of healthcare predictive analytics. The major purpose of this research was to explore the potential applications of big data analytics and machine learning-based approaches in the field of diabetes. In this study, we used the neutrosophic AHP as a feature selection method. The neutrosophic AHP is used to compute the importance of features, then apply the machine learning methods to these features. This study applied logistic regression, support vector machine (SVM), and random forest (RF) to predict the disease of diabetes. 

Keywords: Machine Learning; Diabetes Disease; Neutrosophic, Random Forest; Support Vector Machine; Feature Selection.