A Statistical Framework for Detecting Diabetes Types

Diabetes is a condition defined by the level of hyperglycemia that increases the risk of micro vascular damage. A diabetic patient faces many difficulties such as a reduction in life expectancy, increase in the risk of macro vascular complications, and lower quality of life. The main purpose of this research is to develop a computational system to identify types of diabetes. A Multi Selection Criteria Evaluation technique is used to identify the level of diabetes. In this work, we consider ten common symptoms, which are the causes of metabolic disorder in diabetic patients. We also consider tests that are referred to by doctors for diagnosis of diabetes mellitus. Then, we perform necessary computations to identify the type of diabetes (Type-I or Type-II) by using a K-Nearest Neighbor algorithm, and suggest effective drugs for different types of diabetes. We perform several experiments to show the effectiveness of our framework.