Matthew M. Graham

TL;DR: The performance of Theano is compared against Torch7 and TensorFlow on several machine learning models and recently-introduced functionalities and improvements are discussed.

TL;DR: In this paper, the authors describe a general way to clamp and update the random numbers used in a pseudo-marginal method's unbiased estimator, and obtain more robust Markov chains, which often mix more quickly.

TL;DR: A method for performing efficient MCMC inference in generative models when conditioning on observations of the model output is presented, using a constrained variant of Hamiltonian Monte Carlo which leverages the smooth geometry of the manifold to coherently move between inputs exactly consistent with observations.

TL;DR: In this article, a constrained variant of Hamiltonian Monte Carlo (HMC) is used to integrate a density across the manifold corresponding to the set of inputs consistent with the observed outputs.

TL;DR: In this paper, the authors propose a method for augmenting the Hamiltonian system with an extra continuous temperature control variable which allows the dynamic to bridge between sampling a complex target distribution and a simpler unimodal base distribution, which improves mixing in multimodal targets and allows the normalisation constant of the target distribution to be estimated.