Abstract
The ability to accurately predict the hospitalization duration of an accident and emergency (A&E) patient’s hospital stay is essential for improving patient care, managing healthcare resources, and overall hospital efficiency. This study examined several machine learning prediction models, such as the Decision Tree regression model, Support Vector Regression, and Random Forest regression model, to predict hospitalization for A&E patients. The Random Forest Regression model stands out as the best-fit model and a reliable predictor of hospitalization duration among these models, with a Mean squared error (MSE) value of 11.6079. In this quantitative research-based study, we used a secondary dataset, the MIMIC-III dataset, which has an abundance of patient and medical information and parameters that helped to ensure the authenticity and applicability of the generated predictive models. The results of this study will improve healthcare management by providing an accurate and efficient approach for predicting the length of an A&E patient’s hospital stay.