摘要
本文以1986—2003年全国、江苏省、南京市3个尺度的入境旅游客源为例,基于线性回归模型、移动平均预测模型、指数平滑预测模型以及灰色预测模型,对比研究了各模型在不同空间尺度入境客源预测方面的客观差异,所得结论为:(1)对于不同空间尺度的入境旅游人数预测而言,因不同模型的预测误差差异较大,所以,应选择与尺度特征对应的合适的预测模型。(2)在进行全国尺度的入境旅游人数预测时,可选择以时间为自变量的线性回归预测模型和移动平均预测模型;在进行省级和省级以上尺度的入境旅游人数预测时,可选择移动平均预测模型;在进行市级和市级以上以及更小尺度上的入境旅游人数预测时,可选择指数平滑预测模型。本文研究结论可为不同尺度的旅游客源预测提供有益参考。
The paper, taking inbound tourists in China, Jiangsu Province and Nanjing from 1996 - 2003 as an example, makes a comparative study of the objective difference of different forecast models in different spatial scales. The conclusions obtained are: 1. So far as the forecast of the number of inbound tourists in different spatial scales is concerned, there exist bigger errors in terms of forecast in different models, so it is imperative to select suitable forecast model that corresponds to the scales 2. In forecasting the number of inbound tourists nationwide and above provincial level, it is advisable to select linear return forecast model or movable average forecast model. In forecasting the number of tourists under municipal level, it is advisable to select index smooth forecast model. The study conclusion can provide reference for the forecast of tourists in different scales.
出处
《旅游学刊》
CSSCI
北大核心
2007年第11期17-21,共5页
Tourism Tribune
基金
国家自然科学基金项目成果(编号:40301002)
关键词
尺度
客源预测
预测模型
预测误差
forecast of passenger source
forecast model
forecast