The Time Series Prediction of Tourist Populations Based on ARIMA Model in Xizang

On the basis of exploring the implementation steps of ARIMA time series prediction, mathematical models are introduced such as autocorrelation function, partial autocorrelation function and autoregressive model, moving average model, autoregressive moving average model and the method for determining the three parameters p, d and q of ARIMA. In order to understand the future development tourist populations, it is selected ARIMA model as the prediction experiment in Xizang. This experiment takes the tourist population from 1980 to 2024 as the time series data in Xizang, which are identified ARIMA (1, 1, 0) model by stationarity test, differential processing and model recognition. It is confirmed through both experimental parameters and goodness of fit by model estimation with hypothesis testing and other operations et al., and it also predicts the tourist population of Xizang in the next five years. The experimental results indicate that most of the fitting indicators of the model are sound, which achieved the actual demand for prediction.