Sampling from weights posterior for linear models #38
Description
I was wondering if there is a clean way to sample from the posterior of weights for linear models. Specifically, I think EBLogisticRegression does not return the posterior covariance. Is that correct? I was wondering if there is a reason for that.