Bayesian Neural Networks With Maximum Mean Discrepancy Regularization
arXiv:2003.00952 · doi:10.1016/j.neucom.2021.01.090
Abstract
Bayesian Neural Networks (BNNs) are trained to optimize an entire distribution over their weights instead of a single set, having significant advantages in terms of, e.g., interpretability, multi-task learning, and calibration. Because of the intractability of the resulting optimization problem, most BNNs are either sampled through Monte Carlo methods, or trained by minimizing a suitable Evidence Lower BOund (ELBO) on a variational approximation. In this paper, we propose a variant of the latter, wherein we replace the Kullback-Leibler divergence in the ELBO term with a Maximum Mean Discrepancy (MMD) estimator, inspired by recent work in variational inference. After motivating our proposal based on the properties of the MMD term, we proceed to show a number of empirical advantages of the proposed formulation over the state-of-the-art. In particular, our BNNs achieve higher accuracy on multiple benchmarks, including several image classification tasks. In addition, they are more robust to the selection of a prior over the weights, and they are better calibrated. As a second contribution, we provide a new formulation for estimating the uncertainty on a given prediction, showing it performs in a more robust fashion against adversarial attacks and the injection of noise over their inputs, compared to more classical criteria such as the differential entropy.
References in corpus (8)
- On Calibration of Modern Neural Networks
- Weight Uncertainty in Neural Networks
- Generative Moment Matching Networks
- A Comprehensive guide to Bayesian Convolutional Neural Network with Variational Inference
- On the Convergence of Stochastic Gradient MCMC Algorithms with High-Order Integrators
- Statistical Inference for Generative Models with Maximum Mean Discrepancy
- MMD-Bayes: Robust Bayesian Estimation via Maximum Mean Discrepancy
- How Good is the Bayes Posterior in Deep Neural Networks Really?