Predicting Hospital Admissions using Logistic Regression and Random Forest

Too many patients in the Emergency Departments (EDs) can have negative consequences for patients. There is a need of some innovative methods to prevent overcrowding and improve patient flow. In order to accomplish this we use machine leaning techniques. Some of the existing algorithms to build the predictive models are: (1) logistic regression, (2) decision trees, and (3) reconstructability analysis. In this paper logistic regression and Rain forest technique is used in predicting if a particular patient will get admitted in the ED in his/her next visit. Random forest is a machine learning algorithm. It is an ensemble of weak decision trees in order to form a strong decision tree. It flexible and easy to use and produces accurate result mostly even without hyper-parameter tuning. It is simple and can be used for both classification and regression tasks. So it is a good practice to adopt logistic regression and random forest in predicting a patients admission.