Good Initializations of Variational Bayes for Deep Models

Abstract

Stochastic variational inference is an established way to carry out approximate Bayesian inference 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. The proposed method is extensively validated on regression and classification tasks, including Bayesian DeepNets and ConvNets, showing faster and better convergence compared to alternatives inspired by the literature on initializations for loss minimization.

arXiv comment8 pages of main paper (+3 for references and +6 of supplement material)

Cited by in corpus (5)

• Generalized Variational Inference: Three arguments for deriving new Posteriors
• All You Need is a Good Functional Prior for Bayesian Deep Learning
• Walsh-Hadamard Variational Inference for Bayesian Deep Learning
• Quantifying Model Uncertainty in Inverse Problems via Bayesian Deep Gradient Descent
• Efficient Approximate Inference with Walsh-Hadamard Variational Inference