Predicting Hotel Reservation Cancellation by Using Machine Learning Methods

In order to maximize profit for hotels, occupancy rates must be high. For this reason, hotels should allocate a limited number of their rooms to the right customer at the right time using reservation systems software. However, reservations may be cancelled by the customer for various reasons. Cancellations may result for hotels in loss of income if the right policies are not processed. For this reason, it is very important to estimate reservation cancellations. In this study, the hotel reservation data set consisting of 38,826 records of 5 different hotels were analyzed by machine learning algorithms to estimate the cancellation of future bookings of hotels. In this context, 4 different models were formed in this study by using Random Forest (RF), Support Vector Machines (SVM), k-Nearest Neighbor (kNN) and Decision Tree (C4.5) algorithms and then, performance comparisons were made among these models. The best result was obtained from C4.5 decision tree algorithm with 73% accuracy