Machine Learning Approaches to Classify Diabetes Patients based on Age, Obesity level and Cholesterol level

Abstract

Nowadays finding the root cause of some diseases and their effect on different organs of the human body is challenging rather than treatment of the disease. Diabetes stands first in that category. Diabetes is a condition in which the body is incapable of producing insulin or it is not in a situation to make use of the produced insulin, and sometimes both. It is also called as Diabetes Mellitus. In this paper, we experimented machine learning algorithms to find the impact of age obesity level (O), and cholesterol level (C) of a person on diabetes. We have collected real-time data of 50 patients, where 34 are nondiabetic and 16 are diabetic. Each record consists of AOC of a person along with class label attribute. Experimental results unfold that, there is a significant effect of obesity and cholesterol in diabetic patients. Results are compared using rule-based classifier JRip, probability-based classifier Naïve Bayes, and decision tree based classifiers J48, Random forest. All experiments are conducted using 10 fold cross validations by considering the random blood sugar levels of patients.

Sule Lamido University Journal of Science and Technology (SLUJST) Vol. 3 No. 1&2 [June, 2022], pp. 113-121113Obesity Level ClassificationBased on Decision Tree and Naïve Bayes Classifiers

“…A limited number of publications exist that address using machine learning techniques to classify obesity (De-la-hoz-correa et al, 2019;Priyaa, Sathyapriya, & Arockiam, 2020). Machine learning is being used to examine large amounts of data to identify patterns and relationships that would otherwise go undetected (Cheng et al, 2021;Dugan et al, 2015;Nimmala, 2019). Some studies have applied these methods to build predictive models to understand the obesity problem.…”

“…Before automated classification takes place in this modern technology, the traditional method has been used to analyze the individual parameters and manually classified by domain experts based on the biological and social attributes. Several classification techniques have been applied in text classification such as Decision Tree (DT) classifiers, Naïve Bayes (NB) probabilistic classifiers, Neural Network, K-Nearest Neighbours (KNN), Support Vector Machine (SVM), and Rocchio classifiers (Beunza et al, 2019;Dugan, Mukhopadhyay, Carroll, & Downs, 2015;Nimmala, 2019).…”