旅游客流量预测:基于季节调整的PSO-SVR模型研究  

Study on tourism flow forecast: based on seasonal adjustment'PSO-SVR model

作　　者：翁钢民[1] 李凌雁[1] 

机构地区：[1]燕山大学经济管理学院,河北秦皇岛066004

出　　处：《计算机应用研究》2014年第3期692-695,共4页Application Research of Computers

基　　金：国家社会科学基金资助项目(09BJY087);河北省社会科学基金资助项目(HB12YJ075)

摘　　要：旅游客流量具有明显的非线性和季节性特征,所以采取季节调整方法对样本数据进行预处理,消除季节性的影响,可以提高客流量预测的准确性。同时SVR(支持向量回归机)是一种良好的机器学习方法,非常适合预测研究,辅以PSO(粒子群算法)选取合适的回归参数可以获得更加精确的预测结果。鉴于此,构建一种考虑季节影响的PSO-SVR模型,以北京为例将不同旅游客流量预测方法的拟合优度进行比较。结果显示:季节调整的PSO-SVR模型预测精度明显高于SVR、季节调整的SVR和PSO-SVR模型,该模型是进行旅游客流量预测的有效工具。Tourist industry has a strong seasonal trend. Taking seasonal adjustment method for preprocessing of sample data to eliminate the seasonal influence can improve the accuracy of tourism flow prediction. At the same time, SYR is .a good ma- chine learning method, very suitable for predicting tourism flow, and supplement by the PSO algorithm to choose appropriate parameters, it can obtain more accurate predicting results. In view of this situation, it needed to build a tourism flow predic- tion method which considered seasonal influence and through SVR optimized by the PSO model with Beijing as an eXa^mple for empirical research, and compared the goodness of fit of different prediction methods. The results indicate that the prediction accuracy of seasonally adjusted PSO-SVR model is significantly higher than the SVR model, seasonally adjusted SVR model, the PSO-SVR model. The seasona｜ly adjusted PSO-SVR model is an effective tool of tourism flow prediction.

关 键 词：旅游客流量预测 粒子群算法 支持向量回归机 季节调整 均方差比较 

分 类 号：F590.3[经济管理—旅游管理] TP393[经济管理—产业经济]