Timezone: »

Bayesian dark knowledge
Anoop Korattikara Balan · Vivek Rathod · Kevin Murphy · Max Welling

Thu Dec 10 08:00 AM -- 12:00 PM (PST) @ 210 C #21

We consider the problem of Bayesian parameter estimation for deep neural networks, which is important in problem settings where we may have little data, and/ or where we need accurate posterior predictive densities p(y|x, D), e.g., for applications involving bandits or active learning. One simple approach to this is to use online Monte Carlo methods, such as SGLD (stochastic gradient Langevin dynamics). Unfortunately, such a method needs to store many copies of the parameters (which wastes memory), and needs to make predictions using many versions of the model (which wastes time).We describe a method for “distilling” a Monte Carlo approximation to the posterior predictive density into a more compact form, namely a single deep neural network. We compare to two very recent approaches to Bayesian neural networks, namely an approach based on expectation propagation [HLA15] and an approach based on variational Bayes [BCKW15]. Our method performs better than both of these, is much simpler to implement, and uses less computation at test time.

Author Information

Anoop Korattikara Balan (Google)
Vivek Rathod (Google)
Kevin Murphy (Google)
Max Welling (Microsoft Research AI4Science / University of Amsterdam)

More from the Same Authors