Comparative Analysis of Deep Learning and Machine Learning Models for Predicting Hypo- and Hyperglycemia in Type 2 Diabetes

Diabetes Mellitus, specifically in type 2 diabetes, is a long-term metabolic condition affecting millions throughout the globe. The main challenge in the management of T2D is predicting blood sugar level variations, precisely hypo- and hyperglycemia, to prevent severe complications. This study offers a comparative concept of machine learning and deep predictive models of learning for these conditions with emphasis on Support Vector Machine (SVM) and Long Short-Term Memory (LSTM) networks. This dataset uses patient health records: blood glucose levels, lifestyle parameters, and the other pertinent clinical parameters. SVM, a popular machine learning model, is then compared to LSTM strongly used for managing sequential data. The accuracy, precision, recall, and $\mathbf{F 1}$ score are the focus of the study regarding the prediction performance for these models.