Current browse context:
stat.CO
Change to browse by:
References & Citations
Statistics > Computation
Title: Asymptotically exact inference in differentiable generative models
(Submitted on 25 May 2016 (v1), last revised 2 Mar 2017 (this version, v4))
Abstract: Many generative models can be expressed as a differentiable function of random inputs drawn from some simple probability density. This framework includes both deep generative architectures such as Variational Autoencoders and a large class of procedurally defined simulator models. We present a method for performing efficient MCMC inference in such models when conditioning on observations of the model output. For some models this offers an asymptotically exact inference method where Approximate Bayesian Computation might otherwise be employed. We use the intuition that inference corresponds to integrating a density across the manifold corresponding to the set of inputs consistent with the observed outputs. This motivates the use of a constrained variant of Hamiltonian Monte Carlo which leverages the smooth geometry of the manifold to coherently move between inputs exactly consistent with observations. We validate the method by performing inference tasks in a diverse set of models.
Submission history
From: Matthew Graham [view email][v1] Wed, 25 May 2016 11:10:36 GMT (214kb,D)
[v2] Tue, 12 Jul 2016 20:39:35 GMT (453kb,D)
[v3] Wed, 12 Oct 2016 21:31:22 GMT (465kb,D)
[v4] Thu, 2 Mar 2017 22:07:33 GMT (574kb,D)
Link back to: arXiv, form interface, contact.