Robust Bayesian Analysis
2018-05-14T02:29:41Z (GMT) by
This thesis develops two robust Bayesian inferential methods to handle model misspecification due to the presence of outliers in the data. These new methods are designed to produce robust Bayesian inference based on simple parametric models which may be misspecified. Utilizing robust information from the data, these Bayesian methods are able to produce posterior inference about the fundamental relationship between variables that is robust with respect to outliers.