Predicting hotel booking cancelation with machine learning techniques

Abstract

Purpose The purpose of this study is to develop a model that accurately forecasts hotel room cancelations and further determines the key cancelation drivers. Design/methodology/approach Predictive modeling, specifically the machine learning methods, is used to forecast room cancelations and identify the main cancelation factors. Findings By using three different classification algorithms, this study demonstrates that hotel room cancelation can be accurately predicted using XGBoost, as well as the ensemble method involving Support Vector Machine, Random Forest and XGBoost. Originality/value This study attempted to forecast hotel room cancelations by applying a relatively new method, machine learning. By implementing predictive modeling, one of the most emerging and innovative research methods, this study ultimately provides prediction suggestions in various aspects and levels for hotel management operations.