Predicting Risk of Diabetes in Non-Diabetic Population

Abstract

Diabetes Mellitus is a chronic debilitating disease affecting a major population of the developed and developing countries. The occurrence of type 2 diabetes mellitus (T2DM) is rising rapidly among middle-aged American adults. It has been estimated that the prevalence of diabetes in the United States increased from 7.3% in 1993 to 7.9% by the year 2000, and greater frequencies are forecast for the future Prediction of chronic conditions like DM that have a definable onset can help to guide interventions and health policy development. Prediction of future incidence of this disease will enable adequate fund allocation for delivery of care to be planned. This white paper discusses the approach and statistical models used to predict diabetes mellitus in a population with unknown status for diabetes. The prediction is for at present and at three months’ time frame allowing a practitioner to pick up patients at risk of acquiring diabetes. The problem is modeled as supervised classification problem, training data consisted of all the labelled patients and model accuracy is validated on test data set. Multiple models are built with proper tuning, and their performances are compared. Support Vector Machine and Random Forest have better accuracy compared to other models.