Title: Diabetes detection system using machine learning algorithms

Authors: Salliah Shafi Bhat; Madhina Banu; Gufran Ahmad Ansari; Venkatesan Selvam

Addresses: B.S Abdur Rahman Crescent Institute of Science and Technology, Chennai-600048, India ' B.S Abdur Rahman Crescent Institute of Science and Technology, Chennai-600048, India ' Faculty of Science, Dr. Vishwanath Karth MIT World Peace University (MIT-WPU), Pune – 411 038, India ' B.S Abdur Rahman Crescent Institute of Science and Technology, Chennai-600048, India

Abstract: Diabetes is a major severe disease that affects a lot of people worldwide. Technical advances have rapid impact on many aspects of human life whether it is healthcare profession or any other field. The disorder has an impact on society. Machine learning algorithms (MLA) can aid in predicting the chance of developing diabetes at a young age, and assist in improving diabetes clinical condition. The proposed framework can be used in the healthcare industry for diabetes detection and prediction in North Kashmir. Four MLA have been successfully used in the experimental study, random forest, K-nearest neighbour, support vector machine and naive Bayes, respectively. KNN is the most accurate classifier, with the highest accuracy rate of 97.29% in comparison to the other methods with the balanced dataset. Overall, this study enables us to effectively identify the prevalence and prediction of diabetes.

Keywords: diabetes; machine learning algorithms; MLA; framework; random forest; K-nearest neighbour; KNN; support vector machine; SVM; naive Bayes; diabetes detection system; DDS.

DOI: 10.1504/IJEH.2023.135804

International Journal of Electronic Healthcare, 2023 Vol.13 No.3, pp.231 - 246

Received: 15 Dec 2022
Accepted: 06 Oct 2023
Published online: 05 Jan 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article