摘要
针对参数回归模型受很多函数假设限制和非参数回归模型受"维数灾难"影响问题,构造出半参数线性回归模型。结合半参数线性回归模型的主成分估计和Liu估计方法,提出了半参数线性回归模型的主成分Liu估计方法,分别推导出在无约束条件下、精确约束条件下和随机约束条件下的回归系数的各个数字特征,为进一步深入研究半参数线性回归模型的主成分Liu估计提供了有效的理论基础。
Because the parameter regression model is limited by many function hypotheses and nonparametric regression model is affected by dimensionality disaster,a semiparametric linear regression model is constructed in this paper. By combining of principal component estimation method and Liu estimation method, a new method— the principal component Liu regression estimation method in partially linear regression models,is proposed. Under unconstrained conditions,exact constraints and random constraints,each digital feature of regression coefficient is derived respectively. These properties provides an effective theoretical basis to the study of the principal component Liu estimation of the semi parametric linear regression model.
引文
[1]Massy W F.Principal compoents regression in exploratory statistical research[J].Journal of the American Statistical Association,1965,60(2):234-256.
[2]Hoerl A E,Kennard R W.Ridge regression:Biased estimation for nonorthogonal problems[J].Technomertrics,2000,42(12):55-67.
[3]Fikri Akdeniz,Esra Akdeniz Duran.Liu-type estimator in semiparametric regression models.Journal of Statistical Computation and simulation[J].Journal of Statistical Computation and Simulation,2016,80(8):853-871.
[4]朱丹丹,胡宏昌,张捷.半参数回归模型的广义Liu型估计[J].湖北师范学院学报(自然科学版),2012,32(1):59-63.
[5]黄文焕,戚佳金,黄南天.带线性约束的回归模型参数的Liu估计方法[J].系统科学与数学,2009,29(7):937-946.
[6]韦功鼎.Liu型主成分估计的优良性[D].北京:北京交通大学,2012.
[7]王肖南.半参数可加模型的Liu型估计[D].北京:中央民族大学理学院,2013.