Using Machine Learning to Predict Patient’s Admission Trends in Hospital

In this era of advanced healthcare facilities, where health and technology are immensely integrated, the role of big data analytics is evolving in every aspect of healthcare. The large influx of patients in hospitals is a fatal problem for the hospital management systems. The patient crowding problem can cause potential consequences, including increased mortality rate, unnecessary labor cost, poor customer service, ambulance divergence, and cancelation of equipment. Therefore, it is crucial to utilize machine learning to prevent overcrowding and enhance resource allocation. This research study has used the openly available dataset of patients and a combination of machine learning algorithms such as logistic regression, support vector machine, k-nearest neighbor, and decision tree to analyze patient admission trends for effective decision making. According to the simulation results, the decision tree algorithm achieved the best accuracy with a score of (0.89). While on the parameter of the area under the receiver operating characteristics (AUROC), the logistic regression algorithm performed best with an AUROC score of (0.74), followed by the support vector machine with a score of (0.73). The least performers are KNN and decision tree with AUROC of (0.67) and (0.53), respectively.