Data Based Estimation of Near Future Values of Blood Glucose with K-Nearest Neighborhood Algorithm
S.Shanthi1, Shyamala Bharath2, M.Sujatha3
1Dr.S.Shanthi*, Professors, Department of ECE, Saveetha School of Engineering, SIMATS, Chennai, Tamil Nadu, India.
2Dr. Shyamala Bharath, Professors, Department of ECE, Saveetha School of Engineering, SIMATS, Chennai, Tamil Nadu, India.
3Dr.M.Sujatha, Professors, Department of ECE, Saveetha School of Engineering, SIMATS, Chennai, Tamil Nadu, India.
Manuscript received on September 18, 2019. | Revised Manuscript received on 22 September, 2019. | Manuscript published on October 10, 2019. | PP: 1438-1442 | Volume-8 Issue-12, October 2019. | Retrieval Number: L39451081219/2019©BEIESP | DOI: 10.35940/ijitee.L3945.1081219
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Diabetes Mellitus is a disease due to the disorder of carbohydrate metabolism. This disease affects the people in various ways in short and long time periods. Research is being carried out in pathological, pharmaceutical, clinical, therapeutic aspects to reduce the impediments of Diabetes. Researchers try to foretell the proximateforthcoming blood glucose values so that the patient could be alerted to take appropriate action with the guidance of medical practitioner. The near future prediction of glucose values is very much needed for a successful artificial pancreas. Since the blood glucose values of human depends on many factors like physiological factors, age, body mass index, glucose metabolism, insulin action etc,, prediction of exact values of blood glucose still remains a tough task. The current research work focuses on the application KNN regression for the forecast of nearby blood glucose values. The KNN algorithm uses the feature similarity to predict any new points on the data sets. The proposed work has been tested with 3 different data sets and the results have been analyzed. Promising results have been obtained which could be extended further with real time analysis.
Keywords: Artificial Intelligence, Continuous Glucose Monitoring, Diabetes Mellitus, Machine learning, KNN Algorithm, Prediction.
Scope of the Article: Artificial Intelligence