Weight Uncertainty in Neural Networks

We introduce a new, efficient, principled and backpropagation-compatible algorithm for learning a probability distribution on the weights of a neural network, called Bayes by Backprop. It regularises the weights by minimising a compression cost, known as the variational free energy or the expected lower bound on the marginal likelihood. We show that this principled kind of regularisation yields comparable performance to dropout on MNIST classification. We then demonstrate how the learnt uncertainty in the weights can be used to improve generalisation in non-linear regression problems, and how this weight uncertainty can be used to drive the exploration-exploitation trade-off in reinforcement learning.

# Gaussian Processes and Bayesian Inference # Computer Science # Adversarial Robustness in Machine Learning # Neural Networks and Applications

Charles Blundell, Julien Cornebise, Koray Kavukcuoglu, et al. (2015). "Weight Uncertainty in Neural Networks". arXiv (Cornell University). https://doi.org/10.48550/arxiv.1505.05424