基于半参数模型的插补方法研究

Study on Missing Data Imputation Based on Semi-Parametric Model

数据缺失是在数据收集中普遍存在的现象,因而缺失数据的插补问题就成了数据分析领域的重要命题.常用的插补模型有参数模型、非参数模型和半参数模型,其中关于半参数模型的研究较少,考虑到半参数模型的优越性,文章将半参数回归模型与插补方法相结合,利用最小二乘核估计构建半参数模型,再利用辅助变量对目标变量进行估计来建立缺失数据的插补数据集,进而实现缺失数据的插补,并通过案例探究该方法的实现及其可行性.

Data missing is a common phenomenon. Missing data imputation is an important issue in data analyzing. Parametric model, nonparametric model and semi-parametric model are the common models used to impute missing data. Considering the advantages of semi-parametric model, the essay tries to combine the imputation method with semi-parametric model. Firstly, the essay estimates semi-parametric model through using the least squares kernel estimator. Then covariate vectors are used to estimate the target variable in order to establish a data set that can be used for imputation.