How To Predict Patient Arrival In The Emergency Room

The emergency services of the hospitals are for the most part suffering, facing a constant increase in the number of patients without the financial means to follow. Part of the answer to this worrying situation lies in the optimization of existing resources, for which artificial intelligence techniques are proving promising. In this article, we evaluate this possibility in a concrete way, starting from real data and applying a comparative analysis of 4 state-of-the-art algorithms. An original way of selecting explanatory variables is applied, and also the hyperparameters of the algorithms are chosen with precision. The results show that the most powerful machine learning algorithms currently available are quite capable of making good predictions of the number of patients arriving at the emergency room, provided that they are well applied.