Development of Indian Weighted Diabetic Risk Score (IWDRS) using Machine Learning Techniques for Type-2 Diabetes

Abstract

Undetected pre-diabetes and late diagnosis is a major problem in East Asian countries. Diabetes screening tools such as Diabetes Risk Score (DRS) can effectively help in detecting and preventing the disease among pre-diabetes persons. Several Risk Scores for Type -2 Diabetes have been proposed and being used. In current research, researchers have observed certain issues in the available DRS and advocate the need to address the same. In this study researchers propose a novel Indian Weighted Diabetic Risk Score (IWDRS). Machine Learning Techniques such as distance based clustering with Euclidean distance, k-means algorithm and discretization is used to derive weighted risk score for diabetes risk factors like age, BMI, waist circumference, personal history, family history, diet, physical activity, stress and life quality. Result analysis shows that the proposed approach is better than existing approach in scientific literature.