SLANG: Fast Structured Covariance Approximations for Bayesian Deep Learning with Natural Gradient

Open in new window