A Novel Prediction of Diabetes by Automatic Insulin Therapy Using Machine Learning Algorithm
Year: 2020
Type: article
Abstract: Diabetes mellitus is one of the world’s fast-growing diseases. Differentiation is among the most important decision-making approaches in many real-world problems. In this work, the main objective is to classify the diabetic patient’s data into various levels based upon the values. This will help to assist the required dose which should be provided to the patients through an automatic insulin pump. The efficiency of the different classifiers is measured to assess the reliability of the classification. In this analysis, four common algorithms for machine learning, namely Support Vector Machine (SVM), Naive Bayes (NB), Logistic Regression, Random forest, and decision tree, for the estimation of diabetic mellitus on data from the adult population. Based on the comparison of performance parameters like precision, recall, F1score, and accuracy the algorithms are ranked and selected the best among all. The accuracy value of Logistic Regression is the highest among the other algorithm, therefore Logistic Regression performs best with the patient data in forecasting diabetes mellitus.
Authors: Vinothkumar, B, Ramaswami, M
Institution: Madurai Kamaraj University
Language: English
Subfield: Health Information Management
Field: Health Professions
Domain: Health Sciences
Open Access status: hybrid
OpenAlex ID: W7216851351
