An Intelligent Passenger Flow Prediction Method for Pricing Strategy and Hotel Operations

Hospitality industry plays a crucial role in the development of tourism. Predicting future demand hotel is key step process revenue management. Hotel passenger flow prediction an important guiding formulation pricing and operating strategies. On one hand, can provide decision support for managers effectively avoid waste resources loss caused by customers. other it guarantee priority occupation business opportunities enterprises, which help enterprises adjust their operation strategies reasonably to better adapt market situation. In addition, helpful judge overall condition assess risk level project be built. affected many factors, such as weather, environment, season, holidays, economy, emergencies, has characteristics complex nonlinear fluctuation. The existing predicting methods include linear methods. rely on stability environment time series, so they cannot completely simulate fluctuations flow. Traditional need improve accuracy, are difficult deal with increasing data Based above analysis, this paper constructs deep learning model based Long Short-Term Memory (LSTM) predict number actual monthly arrival bookings. bookings reflect hotel. reduce cancellation or nonarrival due various reasons revenue. experimental part booking dataset resort Portugal from July 1, 2015, August 31, 2017. Artificial neural network (ANN) vector regression (SVR) built benchmark models results show that, compared models, LSTM ability necessary reference hotel's mode arrangement.