基于ARIMA的入境旅游月度过夜人次预测

入境旅游人次预测对旅游管理部门合理配置旅游资源、创新旅游服务模式有很重要的意义。受气候变化、经济发展趋势、文化差异的影响,旅游人次呈现出明显的季节性与非线性特征,管理部门依据经验推断入境旅游过夜人次的难度越来越大。文章采用ARIMA模型,对入境旅游过夜人次进行月度预测更加科学、准确。选取上海市2004-2016年入境旅游月度过夜人次数据为样本,依据AIC、BIC、HQIC最小准则进行参数估计和模型定阶,拟合出入境旅游月度过夜人次预测的最优模型为ARIMA(6,3,0)。运用该模型,对上海市2017年1-12月的入境旅游月度过夜人次进行预测,并将预测值与2017年真实数据比对,其平均绝对误差为3.22%。可见,应用ARIMA对入境旅游月度过夜人次预测有较高信度。

Monthly Inbound Tourism Overnight Arrivals Forecast with ARIMA Model

Forecasting the number of overnight inbound tourists is of great significance for tourism management department to scientifically allocate relevant resources and innovate their service modes. Affected by climate change, economic development trend and social and cultural differences, the number of tourists has obvious seasonal and non-linear characteristics. Therefore, it is increasingly difficult for the management department to forecast the number of overnight inbound tourists based on experience. This paper uses the ARIMA model to make monthly forecast of the number of overnight inbound tourists, which makes the forecasted data more scientific and accurate. The paper, taking the inbound tourism data of Shanghai from 2004 to 2016 as samples, based on the minimum criteria AIC, BIC and HQIC, conducts parameter estimation and model ranking, ARIMA(6,3,0) is selected as the best model to forecast the monthly number of overnight inbound tourists in Shanghai. With the model, the monthly number of overnight inbound tourists in Shanghai from January to December 2017 is predicted. The mean absolute error between the predicted data and the real data in 2017 is 3.22%, which indicates that ARIMA is reliable in forecasting the number of monthly overnight inbound tourists.