A0734
Title: Robust Bayesian analysis based on trimmed mean regression
Authors: Lulu Zhang - The Hong Kong Polytechnic University (PolyU) (Hong Kong) [presenting]
Abstract: The use of Bayesian statistics in the social sciences is becoming increasingly widespread. However, seemingly high entry costs still keep many applied researchers from embracing Bayesian methods. Next to a lack of familiarity with the underlying conceptual foundations, the need to implement statistical models using specific programming languages remains one of the biggest hurdles. We will investigate Bayesian and robust Bayesian estimation of a wide range of parameters of interest in the context of Bayesian nonparametric under a broad class of trimmed mean regression and quantile regression. Dealing with uncertainty regarding the prior, we consider the Dirichlet and provide an explicit form of the resulting robust Bayes estimator.