Paper Title
A Machine Learning Overbooking Algorithm for Enhancing Clinic Efficiency

Abstract
In this paper, we predict patient no-show using a random forest model. To mitigate the effects of patient no-show, we propose a predictive overbooking algorithm as a strategy. The overbooking algorithm was simulated using a single server queue model in which patient appointments for a specific timeslot were scheduled in advanced, based on the probability of no-shows for that day. The focus of this paper was to evaluate our proposed overbooking algorithm compared to the reference scenario of no overbooking, as well as blind overbooking, commonly applied today. The criteria used for evaluation of the three methods was clinic efficiency and clinic profitability. Our findings conclude that predictive overbooking brings about significant improvements in several aspects of clinic efficiency as compared to both no overbooking and blind overbooking. We also observed a significant increase in clinic profits in the simulation of predictive overbooking compared to no overbooking and blind overbooking. Keywords - Patient No-Show, Predictive Overbooking Algorithm, Random Forest Model, Clinic Profitability, Clinic Efficiency, Blind Overbooking, Single Server Queue Model, Overbooking Simulation