\[ \newcommand{\argmin}{\operatorname{argmin}} \newcommand{\sign}{\operatorname{sign}} \newcommand{\diag}[1]{\operatorname{diag}(#1)} \newcommand{\prox}[2]{\operatorname{prox}_{#1}(#2)} \]


0.1 About

“Default Bayesian Analysis with Horseshoe Estimators” is an upcoming paper that addresses the issue of default Bayes analysis for multivariate normals. It is based on the Horseshoe+ prior in “The Horseshoe+ Estimator of Ultra-Sparse Signals”.

0.2 Examples

These examples are currently in development and ultimately for demonstration purposes only. Most examples make use of Stan for estimation.

0.2.1 Normal Mean Problem

Link

0.2.2 Fieller-Creasy (Ratio of Means) Problem

Link

0.2.3 Product of Means Problem

Link

0.2.4 Efron Maximum Problem

Link

0.2.5 Efron Sum of Squares Problem

Link