Bayesian influence diagnostics using normalizing functional Bregman divergence

Ian M Danilevicz, Ricardo S Ehlers

Ideally, any statistical inference should be robust to local influences. Although there are simple ways to check about leverage points in independent and linear problems, more comp…