A theory of data variability in Neural Network Bayesian inference

Javed Lindner, David Dahmen, Michael Krämer +1

Bayesian inference and kernel methods are well established in machine learning. The neural network Gaussian process in particular provides a concept to investigate neural networks…