This paper investigated the problem of variable selection and structure identification in varying coefficient model under robust regression.By using B-spline basis function to approximate the non-parametric part,it proposed a combined penalization procedure to select the significant variables,detect the true structure of the model and estimate the unknown regression coefficients simultaneously.It proved that under certain conditions,the consistency and sparsity of the proposed procedure are valid.In addition,it illustrated finite sample performances of the proposed method through some simulation studies.
关键词
变系数模型/稳健回归/自适应组Lasso/变量选择/稀疏性
Key words
varying coefficient models/robust regression/adaptive group Lasso/Variable selection/sparsity
维普
万方数据