节点文献

基于支持向量机与时间序列组合模型的我国旅游市场预测研究

On Tourism Market Prediction Based on Support Vector Machine and Time Series Combination Model

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 高孝成范英兵肖钧莹

【Author】 Gao Xiaocheng;Fan Yingbing;Xiao Junying;Heihe University;

【机构】 黑河学院

【摘要】 我国国内旅游行业已成为世界上增加速度最快、潜力最大的旅游市场,其在经济发展中占据非常重要的位置。因此,如何在未来中把握和预测国内旅游市场的发展趋势,为旅游企业和管理者做出正确决策成为政府需解决的现实问题。本文选取1994—2019年国内旅人次数据,建立了ARIMA模型、SVM模型和两者的组合模型,ARIMA与SVM模型相比,SVM模型预测的精确度较高,但组合模型的精确度最高,拟合相对误差在0.1%以下。利用组合预测模型得到2020年国内旅游人次6 604.396 6万,国内旅游市场的规模会不断壮大,为政府相关部门制定旅游规划政策等提供支持。

【Abstract】 China domestic tourism industry has become the world’s fastest growing and a most potential tourism market, occupying a very important position in the economic development. The realistic problems that the government must solve at present is to grasp and predict the development trend of domestic tourism market in future and to make correct decisions for tourism enterprises and managers. This paper selects the domestic traveler data from 1994 to 2019, and establishes the ARIMA model, the SVM model and the combined model of the two. Compared with the SVM model, the SVM model has higher prediction accuracy, but the combined model has the highest accuracy. The relative error is below 0.1%. Using the combined forecasting model, we will get 660,043,966 domestic tourist visits in 2020. The scale of the domestic tourism market will continue to grow and provide support for relevant government departments to formulate tourism planning policies.

  • 【文献出处】 黑河学院学报 ,Journal of Heihe University , 编辑部邮箱 ,2021年04期
  • 【分类号】F592
  • 【被引频次】1
  • 【下载频次】461
节点文献中: 

本文链接的文献网络图示:

本文的引文网络