Rethinking Approximate Gaussian Inference in Classification

Open in new window