Substantiating the effect of DXA variables in the prediction of diabetes mellitus using machine learning

Author(s):  
Rohith Haridas ◽  
Gautham Nandakumar ◽  
Girish Srinivasan ◽  
Youngsung Yoo ◽  
Hansuk Kim ◽  
...  
2020 ◽  
Vol 17 (8) ◽  
pp. 3449-3452
Author(s):  
M. S. Roobini ◽  
Y. Sai Satwick ◽  
A. Anil Kumar Reddy ◽  
M. Lakshmi ◽  
D. Deepa ◽  
...  

In today’s world diabetes is the major health challenges in India. It is a group of a syndrome that results in too much sugar in the blood. It is a protracted condition that affects the way the body mechanizes the blood sugar. Prevention and prediction of diabetes mellitus is increasingly gaining interest in medical sciences. The aim is how to predict at an early stage of diabetes using different machine learning techniques. In this paper basically, we use well-known classification that are Decision tree, K-Nearest Neighbors, Support Vector Machine, and Random forest. These classification techniques used with Pima Indians diabetes dataset. Therefore, we predict diabetes at different stage and analyze the performance of different classification techniques. We Also proposed a conceptual model for the prediction of diabetes mellitus using different machine learning techniques. In this paper we also compare the accuracy of the different machine learning techniques to finding the diabetes mellitus at early stage.


2020 ◽  
Vol 8 (5) ◽  
pp. 2376-2381

Today world is extensively affected by endocrine disease Diabetes Mellitus which is commonly known as diabetes. There is a need for an effective model which can predict diabetes and its types at the early stages with accuracy. To improve the accuracy of prediction and to achieve better efficiency, a new Machine Learning based Model (MLM) is proposed. This Machine Learning Model (MLM) has ability to predict the diabetes and its categories as type 1, type 2 and Gestational diabetic with which the patient is suffering from. The proposed Machine Learning Model is innovative for diagnosis of diabetes is more accurate as compared to other existing approaches.This is a novel method from which one can combine power of an expert system with the machine learning environment.


Sign in / Sign up

Export Citation Format

Share Document