Classifier Algorithms and Ensemble Models for Diabetes Mellitus Prediction: A Review

Abstract

Timely disease prediction means a lot to the improvement of the health care services and this will go a long way to assist populaces to avoid unsafe health circumstances before resulting in complex medical situations. Diabetes Mellitus as one of the deadly diseases that is described by hyperglycemia taking place due to defects in insulin secretion which allow an irregular increase in glucose level. Diabetes Mellitus can lead to loss of sight, non-traumatic lower extremity amputation, chronic kidney disease, coronary heart disease, stroke, etc. Hence, prompt diagnosis of the Diabetes Mellitus disease has to pay more attention to in the recent area of research. Presently, there is a great and wide work carried out on Machine Learning with a focus on medical and its application. This paper reviewed recent journalsthat made use of Artificial Intelligencetechniques, different classifiers and ensemble methods to assist in the management of diabetes. Classifiers algorithms such as Naive Bayes, Decision Tree, Artificial Neural Network, Support Vector Machine, K-Nearest Neighbour and Multi-Layer Perceptron. The results from over 1,137most related reviewed journalsreveled that Ensemble Models have the highest average accuracy of 87.09 % in respect to prediction of diabetes mellitus

Early Detection of Diabetes Using Random Forest Algorithm

“…According to the World Health Organization (WHO), diabetes mellitus is a chronic condition that occurs when the pancreas is no longer able to produce enough insulin or when the body is not able to use insulin effectively, which leads to the occurrence of increased levels of glucose (sugar) in the blood, which can cause various health complications [2]. Diabetes can be classified into four types: type 1, type 2, gestational, and prediabetes [3]. The long-term effects of the disease include blindness, kidney failure, amputation, and even death.…”