Exploring an overview of machine learning techniques for identifying diabetes: A review

Abstract

Diabetes is among the most severe long-term complications of metabolic disease, typically arising from higher levels of blood sugar spurred on by either inadequate insulin synthesis or inappropriate insulin use by the body. Numerous health problems, including nerve damage, renal problems, and heart disease, may arise from it. Controlling diabetes and avoiding its complications requires proper treatment, which includes medication and lifestyle modifications. The goal of this review study is to highlight recent developments and problems in the area of diabetes identification by analysing lot of Machine Learning (ML) approaches. Conventional ML-based approaches are among the main categories into which the review divides current methodologies. We explore the techniques, guiding concepts, and noteworthy algorithms for each category, offering a thorough analysis of their advantages as well as suggestions for future research. © 2025 IEEE.