Machine Learning Classifiers to Decrease Diabetic Patients Probability of Hospital Readmission

Diabetes is a chronic condition that affects millions of individuals around the globe. Unfortunately, patients with diabetes have a significant risk of hospital readmission following discharge. Using machine learning classifiers, proposed a strategy to reduce the risk of readmission for 768 diabetes patients. This utilized a dataset containing patient demographic information, medical history, and laboratory results which are collected from Pima India Dataset. To predict the risk of readmission, and train several Machine Learning models, including Logistic Regression, Random Forest, and Support Vector Machines. The research indicates that the Random Forest model has the maximum degree of precision, at 97%. Age, glucose level, Hba1c levels, and comorbidities have been identified as the most significant factors contributing to the risk of readmission. On the basis of these findings, this recommends a personalized approach to diabetes management that takes individual patient characteristics and risk factors into account. This strategy could reduce the likelihood of readmission for diabetic patients and enhance their overall health.