Gibbs posterior concentration rates under sub-exponential type losses

Nicholas Syring, Ryan Martin

Bayesian posterior distributions are widely used for inference, but their dependence on a statistical model creates some challenges. In particular, there may be lots of nuisance pa…