Hybrid long short-term memory and decision tree model for optimizing patient volume predictions in emergency departments

Abstract

In this study, we address critical operational inefficiencies in emergency departments (EDs) by developing a hybrid predictive model that integrates long short-term memory (LSTM) networks with decision trees (DT). This model significantly enhances the prediction of patient volumes, a key factor in reducing wait times, optimizing resource allocation, and improving overall service quality in hospitals. By accurately forecasting the number of incoming patients, our model facilitates the efficient distribution of both human and material resources, tailored specifically to anticipated demand. Furthermore, this predictive accuracy ensures that EDs can maintain high service standards even during peak times, ultimately leading to better patient outcomes and more effective use of healthcare facilities. This paper demonstrates how advanced data analytics can be leveraged to solve some of the most pressing challenges faced by emergency medical services today.