Random forest and genetic algorithm united with hyperparameter for diabetes prediction by using WBSMOTE, wrapper approach
by A. Usha Nandhini; K. Dharmarajan
International Journal of System of Systems Engineering (IJSSE), Vol. 13, No. 2, 2023

Abstract: Food is converted into energy by the human body, but diabetes develops when insulin stops working properly and glucose remains in the bloodstream. Heart disease, stroke, renal failure, blindness, nerve damage, gum disease, and even amputations can all be caused by hyperglycemia, or high blood sugar. In recent years, machine learning has made great strides, and its usage has improved numerous areas of healthcare. This research aimed to construct a model that could accurately predict a person's likelihood of developing diabetes. In this paper, we focus on preprocessing techniques and the problem of data imbalance. In this research, diabetes diagnosis was accomplished using the random forest classifier (RFC), WBSMOTE, and the wrapping method. Accuracy in the RFC was improved when evolutionary algorithms were used with the hyperparameter optimisation technique. The UCI machine learning repository's PIMA Indian Diabetes (PIDD) dataset was used for the tests. The outcomes demonstrated that the suggested method outperformed with a maximum accuracy of 93%.

Online publication date: Thu, 01-Jun-2023

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of System of Systems Engineering (IJSSE):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com