Detecting diabetes in an ensemble model using a unique PSO-GWO hybrid approach to hyperparameter optimization

Diabetes is a chronic medical condition that disrupts the body's normal blood sugar levels. It is essential to detect this disease at an early stage in order to prevent organ and tissue injury. This study focuses on diagnosing diabetes by leveraging ensemble learning methods, which involve combining various machine learning techniques. The goal is to create an ensemble learning model that achieves the best classification performance by employing different classifiers and combining techniques. The study explores boosting, bagging, voting, and stacking ensemble learning methods, while also introducing an approach called PSO-GWO (Particle Swarm Optimization and Grey Wolf Optimization) hybrid method for optimizing the model's hyperparameters. The model consisting of combining various classifiers in the stacking ensemble learning method provided the highest classification performance in diagnosing diabetes. The 5-fold cross-validation method is used in the study. Within the scope of the study, the highest accuracy with (98.10%) is obtained with the random forest classifier. The results of the study are presented in comparison with other studies in the literature. These findings contribute to the field of diabetes diagnosis and highlight the potential for developing more accurate and reliable diagnostic systems in the future.