Machine Learning-Based Classification Models for Diagnosis of Diabetes

Abstract

Introduction: The goal of this study is to expand the diabetes decision-making framework through the advancement of computational intelligence. Several artificial network and machine- learning-based methods have been developed and validated, most of which are based on the Pima Indian dataset. So far, no method has reached an accuracy of 99-100%. Various tools such as Machine Learning (ML) and Data Mining are used for the correct identification of diabetes. These tools improve the diagnostic process associated with T2DM. Diabetes mellitus type 2 (DMT2) is a major problem in several developing countries, and its early diagnosis can save several people’s lives. Accordingly, we have to build a structure that diagnoses type 2 diabetes. This paper proposes a fuzzy expert system that uses the Mamdani fuzzy inference structure (MFIS) to diagnose type 2 diabetes accurately. The proposed research work has been created using a variety of machine learning algorithms such as J48 Decision-tree (DT), Multilayer perceptron (MLP), Support-vector-machine (SVM), Naive-Bayes (NB), Fusion, and Mixed fusion-based. Actual data from the UCI machine learning datasets are used to validate the advanced Fuzzy expert system (FES) and machine learning algorithms. Objective: A review of recent advances in machine learning-based classification models for diabetes diagnosis is presented in this survey paper. Methods: This paper compares modified fusion processes to fundamental models such as radial basis function, K-nearest neighbor, support vector machine, J48, logistic regression, classification, regression trees, etc., for diagnosing type 2 diabetes. Results: Figs. 3 and 4 show the results for each classifier based on prediction accuracy. Conclusion: The fuzzy expert system is the best among its rival classifiers. SVM performs very poorly with a very low true positive rate, i.e., a very high number of positive cases misclassified as (non-diabetic) negative. Based on the evaluation, it is clear that the fuzzy expert system has the highest precision value. However, J48 is the least accurate classifier. Compared to the other classifiers listed in the testing section, it has the greatest number of false positives. The results show that the fuzzy expert system has the uppermost cost for both precision and recall. Thus, it has the uppermost value for F-measure in the training and testing datasets. J48 is considered the secondbest classifier for the training dataset, whereas Naïve Bayes comes in the second rank in the testing dataset.