通货膨胀的非参数回归模型  

Nonparametric Regression Model of Inflation

作　　者：张德生[1] 张小静[1] 武新乾[2] 

机构地区：[1]西安理工大学理学院,西安710054 [2]西北工业大学理学院,西安710072

出　　处：《西安工业大学学报》2007年第1期95-98,共4页Journal of Xi’an Technological University

摘　　要：通货膨胀对一个国家经济增长和人民生活的影响是不言而喻的,居民消费价格指数是衡量通货膨胀的重要指标.讨论了非参数回归模型的多项式样条估计方法后,运用我国居民消费价格指数和商品出口额的数据,建立了我国通货膨胀的非参数回归模型,并和线性回归模型的最小二乘估计及非参数局部线性回归估计的结果进行比较,结果表明,在估计和预测上,多项式样条方法都优于线性最小二乘估计和局部线性回归估计,能够更好的反映两者之间的关系.Inflation, for which consumer price index is an important index to measure, has a strong impact on economic development and people＇s life. The polynomial spline estimation for nonparametric regression model is discussed and the nonparametric regression model of inflation is established using the data for CPI and commodity exports. The ploynomial spline esimation is compared with the least squares estimation of linear regression model and the local linear estimation. Results show that nonparametric regression model is better than that of linear regression, and local linear estimation, on both estimation and forecast, and represent well the relations between the two items.

关 键 词：通货膨胀 线性回归 局部线性回归 多项式样条估计 预测 

分 类 号：F822.5[经济管理—财政学] F224