Poster
A Simple Baseline for Bayesian Uncertainty in Deep Learning
Wesley J Maddox · Pavel Izmailov · Timur Garipov · Dmitry Vetrov · Andrew Gordon Wilson

Tue Dec 10th 10:45 AM -- 12:45 PM @ East Exhibition Hall B + C #146

We propose SWA-Gaussian (SWAG), a simple, scalable, and general purpose approach for uncertainty representation and calibration in deep learning. Stochastic Weight Averaging (SWA), which computes the first moment of stochastic gradient descent (SGD) iterates with a modified learning rate schedule, has recently been shown to improve generalization in deep learning. With SWAG, we fit a Gaussian using the SWA solution as the first moment and a low rank plus diagonal covariance also derived from the SGD iterates, forming an approximate posterior distribution over neural network weights; we then sample from this Gaussian distribution to perform Bayesian model averaging. We empirically find that SWAG approximates the shape of the true posterior, in accordance with results describing the stationary distribution of SGD iterates. Moreover, we demonstrate that SWAG performs well on a wide variety of tasks, including out of sample detection, calibration, and transfer learning, in comparison to many popular alternatives including variational inference, MC dropout, KFAC Laplace, and temperature scaling.

Author Information

Wesley J Maddox (New York University)
Pavel Izmailov (New York University)
Timur Garipov (MIT CSAIL)
Dmitry Vetrov (Higher School of Economics, Samsung AI Center, Moscow)
Andrew Gordon Wilson (New York University)

More from the Same Authors