旅游客源预测模型及其对比

Forecast Models and Their Comparison on International Tourism Passenger Source of China

以中国 1978— 2 0 0 1年入境客源为例 ,定量分析线性回归模型、移动平均预测模型、指数平滑模型以及灰色预测模型的应用及其差异问题 ,得出以下结论 :1)对于 2 1a、8a、5a序列客源数据 ,线性回归预测模型的绝对误差最大 ;2 )随着序列数据从 2 1a向 8a、5a变化 ,指数平滑模型绝对误差整体相对变大 ,而灰色模型绝对误差整体相对变小 ;3)在进行我国入境客源预测时 ,如果序列较长 (8a或以上 ) ,可选择指数平滑预测模型 ,如果序列较短 (4～ 7a) ,可选择灰色预测模型 。

It is useful to exactly forecast the tourism passenger source in tourism layout and management.The linear regression model is often used in passenger source in tourism layout.But its precision is worth discussing.Quantitative analysis is made to the four forecast models in a case study of the international tourism passenger source of China.The research suggests that the linear regression model has the lowest precision in the four models,the Grey model of GM(1,1)is suitable for the data of 4～7 years,and the exponent model fits for the data of longer series.