An Early Diabetes Diseases Prediction Using Machine Learning With Optimal Features Selection

Diabetics have elevated blood glucose levels for a lengthy period of time, either due to insufficient insulin synthesis or a lack of effective insulin response by the body's cells. Long-term harm, breakdown, and collapse of diverse organs, including the eyes, kidneys, nerves, heart, and veins,...

Full description

Saved in:

Bibliographic Details
Published in:	2023 International Conference on Research Methodologies in Knowledge Management, Artificial Intelligence and Telecommunication Engineering (RMKMATE) pp. 1 - 7
Main Authors:	Kasinathan, Prabakaran, Sangeetha, M, Balamurugan, M, Aminabee, Shaik, Beema Rao, B Dean, Reddy, Pundru Chandra Shaker
Format:	Conference Proceeding
Language:	English
Published:	IEEE 01-11-2023
Subjects:	Diabetes feature selection Insulin Machine learning Maximum likelihood estimation prediction Prediction algorithms Predictive analytics Predictive models Random Forest Veins
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Description
Summary:	Diabetics have elevated blood glucose levels for a lengthy period of time, either due to insufficient insulin synthesis or a lack of effective insulin response by the body's cells. Long-term harm, breakdown, and collapse of diverse organs, including the eyes, kidneys, nerves, heart, and veins, are associated with the persistent hyperglycemia of diabetes. The focus of this study is on utilizing important features, developing a prediction algorithm by means of Machine-learning(ML), and identifying the best classifier in order to obtain the most accurate results when compared to clinical outcomes. The proposed approach utilizes Predictive analysis with the goal of zeroing in on the variables that are crucial for early diagnosis of Diabetes Miletus. There were a variety of methods used, including random-forest(RF), extreme-gradient-boost(EDB), logistic-regression(LR), and weighted-ensemble-models(WSM). When compared to severe gradient boost (0.93), logistic regression (0.92), and weighted ensemble model (0.87), the RF performed admirably in anticipating uncontrolled diabetes. The study also generalizes the process of picking the best features from the dataset to boost classification precision. These clinical features can be utilized in conjunction with machine learning methods for diabetes control prediction.
DOI:	10.1109/RMKMATE59243.2023.10369382