On the optimization and pruning for Bayesian deep learning
This addresses uncertainty quantification and model pruning in deep learning, offering a one-shot optimization and pruning approach, though it appears incremental as it builds on existing variational and MCMC techniques.
The paper tackles the computational intractability of Bayesian deep learning by proposing an adaptive variational Bayesian algorithm and an EM-MCMC method with spike-and-slab priors, achieving state-of-the-art performance on CIFAR-10, CIFAR-100, and ImageNet datasets.
The goal of Bayesian deep learning is to provide uncertainty quantification via the posterior distribution. However, exact inference over the weight space is computationally intractable due to the ultra-high dimensions of the neural network. Variational inference (VI) is a promising approach, but naive application on weight space does not scale well and often underperform on predictive accuracy. In this paper, we propose a new adaptive variational Bayesian algorithm to train neural networks on weight space that achieves high predictive accuracy. By showing that there is an equivalence to Stochastic Gradient Hamiltonian Monte Carlo(SGHMC) with preconditioning matrix, we then propose an MCMC within EM algorithm, which incorporates the spike-and-slab prior to capture the sparsity of the neural network. The EM-MCMC algorithm allows us to perform optimization and model pruning within one-shot. We evaluate our methods on CIFAR-10, CIFAR-100 and ImageNet datasets, and demonstrate that our dense model can reach the state-of-the-art performance and our sparse model perform very well compared to previously proposed pruning schemes.