部分线性模型在贵州地区旅游外汇收入分析与预测中的应用  

Application of Partial Linear Model in the Analysis and Forecast of Tourism Foreign Exchange Income in Guizhou

作　　者：柏庭引 

机构地区：[1]贵州大学数学与统计学院,贵州 贵阳

出　　处：《应用数学进展》2022年第10期6889-6896,共8页Advances in Applied Mathematics

摘　　要：部分线性模型既具有参数模型的灵活性,又能避免维数灾难问题,因此本文考虑用部分线性模型研究贵州地区旅游外汇收入。首先对贵州省2006~2019年旅游外汇收入数据用主成分分析进行降维,用处理后的数据拟合部分线性模型,用拟合后的模型和多元线性回归模型分别对贵州2020年旅游外汇收入值进行预测并分别计算预测值与真实值间的绝对预测误差。部分线性模型所得的绝对预测误差为1.5%,而逐步线性回归方法的绝对预测误差为13.8%。结果表明,本文所用方法优于多元线性回归模型。Partially linear model not only has the flexibility of parameter model, but also can avoid dimen-sional disaster problem. Therefore, this paper considers using partially linear model to study the foreign exchange income of tourism in Guizhou. Firstly, the dimensionality of tourism foreign ex-change income data of Guizhou Province from 2006 to 2019 is reduced by principal component analysis. The processed data are used to fit a partial linear model. The fitted model and multiple linear regression model are used to predict the tourism foreign exchange income of Guizhou Prov-ince in 2020, and the absolute prediction error between the predicted value and the real value is calculated. The absolute prediction error of partially linear model is 1.5%, while the absolute pre-diction error of stepwise linear regression method is 13.8%. The results show that the method is superior to the multiple linear regression model.

关 键 词：贵州旅游外汇收入 主成分分析 部分线性模型 预测 

分 类 号：F592[经济管理—旅游管理]