Timezone: »

 
Poster
Efficient Low Rank Gaussian Variational Inference for Neural Networks
Marcin Tomczak · Siddharth Swaroop · Richard Turner

Thu Dec 10 09:00 AM -- 11:00 AM (PST) @ Poster Session 5 #1479
in Poster Session 6 »

Bayesian neural networks are enjoying a renaissance driven in part by recent advances in variational inference (VI). The most common form of VI employs a fully factorized or mean-field distribution, but this is known to suffer from several pathologies, especially as we expect posterior distributions with highly correlated parameters. Current algorithms that capture these correlations with a Gaussian approximating family are difficult to scale to large models due to computational costs and high variance of gradient updates. By using a new form of the reparametrization trick, we derive a computationally efficient algorithm for performing VI with a Gaussian family with a low-rank plus diagonal covariance structure. We scale to deep feed-forward and convolutional architectures. We find that adding low-rank terms to parametrized diagonal covariance does not improve predictive performance except on small networks, but low-rank terms added to a constant diagonal covariance improves performance on small and large-scale network architectures.

Keywords: Probabilistic Methods -> Latent Variable Models · Probabilistic Methods -> Topic Models

Author Information

Marcin Tomczak (University of Cambridge)
Siddharth Swaroop (University of Cambridge)
Richard Turner (University of Cambridge)

More from the Same Authors