基于多种模型的福建省入境游客量预测

准确的旅游需求预测可以为政府和产业部门的旅游决策提供重要信息。根据福建省1985-2010年的26年入境游客量统计数据,利用移动平均、指数平滑、GM（1,1）、ARIMA（1,1,1）和GM-ARIMA 5种模型对未来5年入境游客量进行尝试性预测。通过对拟合过程的比较,发现以上5种模型均适用于福建省入境游客量的预测,但以GM-ARIMA组合模型拟合效果最佳,二次移动平均模型拟合效果最差。预测结果表明,至2015年福建省入境游客量大约可达520～580万人次。另外,研究结果支持目前国际上普遍认同的组合模型可以有效提高预测精度的观点。

Forecasts of Inbound Tourist Arrivals using Multiple Models in Fujian Province

Accurate forecasts of tourism demand can provide crucial information for tourism policymaking in both the government and business sectors.Based on statistical data during the period of 1985-2010 available in official yearbooks,five forecasting models,including moving average,exponential smoothing,GM（1,1）,ARIMA（1,1,1）,GM-ARIMA combination techniques,were constructed and employed to predict the inbound tourist arrivals for the next 5 years in Fujian Province.The analyses indicate that all the 5 models are appropriate for forecasting inbound tourist arrivals in Fujian Province,and the GM-ARIMA combination model exhibits the best performance,while the moving average method the worst.The results show that the inbound tourist arrivals in Fujian will reach 520-580 million by 2015.The research results also support the general standpoint over the world that the combination model can effectively improve the forecasting accuracy.