Julyan Arbel (INRIA Grenoble - Rhone-Alpes)

will speak on

Understanding priors in Bayesian neural networks at the unit level

Time: 12:00PM
Date: Mon 23rd November 2020
Location: Online [map]

Abstract: We investigate deep Bayesian neural networks with Gaussian weight priors and a class of ReLU-like nonlinearities. Bayesian neural networks with Gaussian priors are well known to induce an L2, 'weight decay', regularization. Our results characterize a more intricate regularization effect at the level of the unit activations. Our main result establishes that the induced prior distribution on the units before and after activation becomes increasingly heavy-tailed with the depth of the layer. We show that first layer units are Gaussian, second layer units are sub-exponential, and units in deeper layers are characterized by sub-Weibull distributions. Our results provide new theoretical insight on deep Bayesian neural networks, which we corroborate with simulation experiments.

(This talk is part of the Statistics and Actuarial Science series.)

PDF notice

Return to all seminars