Using Machine Learning-based SMOTE Analysis with the Light GBM Classification Method to Classify Diabetic Patients

Diabetes is a prevalent and enduring metabolic condition that impacts a significant global population. The timely identification and categorization of this ailment are of paramount importance in order to provide optimal control and therapeutic interventions. Machine learning algorithms have shown encouraging outcomes in the classification of diabetes using patient data. The gradient boosting framework known as LightGBM is often favoured for this particular job because to its notable efficiency and high level of predicted accuracy. Moreover, the Synthetic Minority Over-sampling Technique (SMOTE) may be used to mitigate the issue of class imbalance in the diabetes dataset, a prevalent challenge encountered in the study of medical data. The primary objective of this work is to use the LightGBM classifier for the purpose of classifying the diabetes condition. The objective will be accomplished by the use of SMOTE Analysis, which aims to tackle the challenge of imbalanced data. Subsequently, an examination of both balanced and unbalanced datasets will be conducted, leading to the attainment of a 72 percent accuracy rate. Future researchers in this specific field of inquiry will benefit from the significant level of accuracy shown in this study.