A comparative study of classification algorithms for early detection of diabetes

Abstract

Diabetes is a chronic disease that affects millions of people worldwide. Early detection of diabetes is crucial for preventing or delaying the onset of its associated complications. In this study, in collaboration with Unidade Local de Saúde do Alto Minho (ULSAM), we conducted a comprehensive comparison of various classification algorithms for the early detection of diabetes. We collected and pre-processed a dataset of patient records, containing personal information, associated medical problems and drugs. The dataset was divided into training and testing sets and used to train and evaluate several popular classification algorithms. The results of our study revealed that the Multilayer Perception (MLP), Gradient Boost Machine (GBM) and Random Forest (RF) algorithms had the highest overall performance, closely followed by Support Vector Machines. These findings demonstrate the potential of these algorithms for use in the early detection of diabetes and suggest that further research is needed to refine and optimize these models for clinical use.