From Deep Additive Kernel Learning to Last-Layer Bayesian Neural Networks via Induced Prior Approximation

Open in new window