2020/04/10 by Carlos Lamarche, Lamarche, Carlos, T.E. Parker +1
Computer Science · Mathematics · #Bayesian Methods and Mixture Models #Econometrics (econ.EM) #FOS: Economics and business #Statistical Methods and Bayesian Inference #Statistical Methods and Inference
paper · pdf · doi:10.48550/arxiv.2004.05127
openalex publication_date 2020/04/10 · openalex created_date 2022/07/26 · openalex updated_date 2026/07/28
The existing theory of penalized quantile regression for longitudinal data\nhas focused primarily on point estimation. In this work, we investigate\nstatistical inference. We propose a wild residual bootstrap procedure and show\nthat it is asymptotically valid for approximating the distribution of the\npenalized estimator. The model puts no restrictions on individual effects, and\nthe estimator achieves consistency by letting the shrinkage decay in importance\nasymptotically. The new method is easy to implement and simulation studies show\nthat it has accurate small sample behavior in comparison with existing\nprocedures. Finally, we illustrate the new approach using U.S. Census data to\nestimate a model that includes more than eighty thousand parameters.\n