Predicting Diabetes in Libya Using AI: A case Study

Abstract

This study investigates the application of artificial intelligence (AI) techniques for predicting diabetes in Libya, a region with limited healthcare data. Utilizing a dataset from the Diabetes and Endocrinology Clinic in Al-Marj, Libya, we examined records of 2,009 patients between April 2019 and May 2023. Our methodology involved applying various AI algorithms, including logistic regression, decision trees, random forests, support vector machines (SVMs), and neural networks, to predict diabetes outcomes. These algorithms were assessed using metrics like precision, recall, F1 score, and the area under the receiver operating characteristic (ROC) curve. The findings indicate a promising potential of AI in forecasting diabetes, particularly when analyzing factors such as fasting blood sugar levels, HbA1c levels, hypertension, heart disease, age, and gender. The majority of the algorithms demonstrated high accuracy, suggesting their utility in enhancing healthcare outcomes in Libya. This research not only provides insights into the effectiveness of AI in diabetes prediction but also underscores the importance of such technologies in regions with scarce health data. It opens pathways for further exploration in the use of AI for healthcare improvement in similar contexts.