Mathematical model of tourism demand forecast

The simplest method is to establish a measurement model, which can be cross-sectional modeling as annual forecast or time series modeling as monthly forecast.

The former can be considered as: selecting several tourist areas may determine the variable indicators of the number of tourists, such as the level of national tourist attractions, the maintenance cost of tourist areas, the area of tourist areas, the tickets for tourist attractions and so on. Then select a series of sample data of tourist attractions, make a single equation cross-sectional multiple regression, and finally substitute the variable values of tourist attractions into the estimation equation according to your investigation, and you can predict the number of tourists.

The latter can be considered as follows: establish a monthly autoregressive moving average (ARMA) model about the number of tourists, add some exogenous variables, such as ticket price and seasonal factors, etc. when necessary, and then determine the lag order P and Q through time series test to ensure the stability of the residual term, which can be used as a prediction model for the next month.