Uncertainty
Sampling for Inference in Probabilistic Models with Fast Bayesian Quadrature
Tom Gunter, Michael A. Osborne, Roman Garnett, Philipp Hennig, Stephen J. Roberts
Neural Information Processing Systems http://nips.cc/
A Polynomial Time Algorithm for Log-Concave Maximum Likelihood via Locally Exponential Families
Brian Axelrod, Ilias Diakonikolas, Alistair Stewart, Anastasios Sidiropoulos, Gregory Valiant
Neural Information Processing Systems http://nips.cc/
Variational Bayesian Monte Carlo with Noisy Likelihoods -- Supplementary Material Luigi Acerbi
Now, since as mentioned above the KL divergence is a non-negative quantity, from Eq. S2 we have
Exponential Family Estimation via Adversarial Dynamics Embedding
Bo Dai, Zhen Liu, Hanjun Dai, Niao He, Arthur Gretton, Le Song, Dale Schuurmans
Can a simple yet effective algorithm be developed for estimating general exponential family distributions? There has been extensive prior work addressing this question.
Posterior Re-calibration for Imbalanced Datasets
Neural Networks can perform poorly when the training label distribution is heavily imbalanced, as well as when the testing data differs from the training distribution. In order to deal with shift in the testing label distribution, which imbalance causes, we motivate the problem from the perspective of an optimal Bayes classifier and derive a post-training prior rebalancing technique that can be solved through a KL-divergence based optimization. This method allows a flexible post-training hyper-parameter to be efficiently tuned on a validation set and effectively modify the classifier margin to deal with this imbalance. We further combine this method with existing likelihood shift methods, re-interpreting them from the same Bayesian perspective, and demonstrating that our method can deal with both problems in a unified way.
Export Reviews, Discussions, Author Feedback and Meta-Reviews
First provide a summary of the paper, and then address the following criteria: Quality, clarity, originality and significance. Summary: The paper is motivated by a recent Plos CB paper by Makin et al 2013, which used numerical simulations of a particular latent variable model that (in this model), performing density estimation of sensory population activity yielded latent variables which retained all the information about the underlying stimulus-variable. In the present paper, the authors ask the question of whether these results hold more gernally. They prove mathematically (assuming some sufficient conditions) that this results holds for a range of models, but also provide counter-examples that it does not always hold. Quality: Results seem sound, but I did not check the proof very rigorously.
Sensory Integration and Density Estimation
Joseph G. Makin, Philip N. Sabes
Neural Information Processing Systems http://nips.cc/