Nonparametric Quantile Regression for Homogeneity Pursuit in Panel Data Models

Many panel data have the latent subgroup effect on individuals, and it is important to correctly identify these groups since efficiency of resulting estimators can be improved significantly by pooling information individuals within each group. However, currently assumed parametric semiparametric relationship between response predictors may misspecified, which leads a wrong grouping result, nonparametric approach hence considered avoid such mistakes. Moreover, depend in different ways at various quantile levels, corresponding structure also vary. To tackle problems, this article proposes regression method for homogeneity pursuit models with individual effects, pairwise fused penalty used automatically select number groups. The asymptotic properties are established, an ADMM algorithm developed. finite sample performance evaluated simulation experiments, usefulness proposed methodology further illustrated empirical example.