基于分形自回归模型的广东省旅游人数预测

Predicting Tourist Number of Guangdong Province Based on Fractal Auto-Regressive Model

摘要: 对旅游人数进行有效预测，有利于政府及相关部门及时采取应对客流量的措施，同时也为当地旅游业及经济发展、规划、管理提供参考意见。本文采用2001~2016年广东省接待游客人数的月度数据，利用其具有的长记忆性，建立ARFIMA (分型差分自回归)模型，并对2017年12个月的来广旅游人数分别进行预测，比较真实值与预测值，结果表明预测效果较好。

Abstract: An effective prediction on tourist number is good for government and related departments taking steps timely to cope with passenger flows. At the same time, it can provide proposal for development, planning and management of local economy. This article adopts monthly data of tourist number coming to Guangdong Province from 2001 to 2017, takes advantage of its long memory, establishes ARFIMA model, and separately predicts twelve months’ tourist numbers in 2017. The result shows that the prediction is effective by comparing true value with predicted value.