Answers are grounded in this paper and cite their sources.
This research demonstrates improved robustness against outliers in high-dimensional settings using Bayesian expectile regression methods.
Compared with ordinary least squares (LS) estimation, expectile regression (ER) is more robust to heavy‐tailed errors or outliers in the response variable. However, ER only considers single expectile information and does not determine whether the selected expectile is appropriate. To overcome this problem, we propose a weighted composite expectile regression (WCER) method, which can effectively resist the occurrence of heavy‐tailed errors or outliers. This article studies weighted composite sparse linear ER under high‐dimensional conditions from a Bayesian perspective and extends the linear model to the Tobit model. The benefit of the Bayesian hierarchical framework is that the weights of each component in the composite model can be treated as open parameters, which can be automatically estimated using Markov Chain Monte Carlo (MCMC) sampling. Finally, simulation and empirical analysis illustrate that this method is superior to the single expectile method.
No takes yet. Share an insight, caveat, or question.
Ji et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: