An evaluation framework for diabetes prediction techniques using machine learning

Abstract

Diabetes affects a large segment of society and does not discriminate based on age. Children, young people, or the elderly may be affected by it. By detecting the disease early, clinicians can help patients recover or at least control it. Models based on machine learning algorithms have been proposed by researchers in the field of artificial intelligence to predict disease and determine its type. The purpose of this study was to propose a framework for evaluating studies related to diabetes detection and identification. To develop the proposed model, a systematic review of studies related to the topic was conducted. After proposing and evaluating the framework, 54 relevant studies were evaluated and results inspired by it were drawn.