Good Initializations of Variational Bayes for Deep Models


While there have been effective proposals for good initializations for loss minimization in deep learning, far less attention has been devoted to the issue of initialization of stochastic variational inference. We address this by proposing a novel layer-wise initialization strategy based on Bayesian linear models.

13 Jun 2019
Long Beach Convention & Entertainment Center
300 E Ocean Blvd, Long Beach, CA, United States
Simone Rossi
PhD Student in Bayesian Machine Learning

My research interests include Variational Inference, Bayesian Deep Learning and Bayesian Uncertainty Estimation