Deep intelligent predictive model for the identification of diabetes

Abstract

<abstract> <p>Diabetes mellitus is a severe, chronic disease that occurs when blood glucose levels rise above certain limits. Many complications arise if diabetes remains untreated and unidentified. Early prediction of diabetes is the most high-quality way to forestall and manipulate diabetes and its complications. With the rising incidence of diabetes, machine learning and deep learning algorithms have been increasingly used to predict diabetes and its complications due to their capacity to care for massive and complicated facts sets. This research aims to develop an intelligent computational model that can accurately predict the probability of diabetes in patients at an early stage. The proposed predictor employs hybrid pseudo-K-tuple nucleotide composition (PseKNC) for sequence formulation, an unsupervised principal component analysis (PCA) algorithm for discriminant feature selection, and a deep neural network (DNN) as a classifier. The experimental results show that the proposed technique can perform better on benchmark datasets. Furthermore, overall assessment performance compared to existing predictors indicated that our predictor outperformed the cutting-edge predictors using 10-fold cross validation. It is anticipated that the proposed model could be a beneficial tool for diabetes diagnosis and precision medicine.</p> </abstract>