We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

stat.ML

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Machine Learning

Title: Optimization of Annealed Importance Sampling Hyperparameters

Abstract: Annealed Importance Sampling (AIS) is a popular algorithm used to estimates the intractable marginal likelihood of deep generative models. Although AIS is guaranteed to provide unbiased estimate for any set of hyperparameters, the common implementations rely on simple heuristics such as the geometric average bridging distributions between initial and the target distribution which affect the estimation performance when the computation budget is limited. In order to reduce the number of sampling iterations, we present a parameteric AIS process with flexible intermediary distributions defined by a residual density with respect to the geometric mean path. Our method allows parameter sharing between annealing distributions, the use of fix linear schedule for discretization and amortization of hyperparameter selection in latent variable models. We assess the performance of Optimized-Path AIS for marginal likelihood estimation of deep generative models and compare it to compare it to more computationally intensive AIS.
Subjects: Machine Learning (stat.ML); Machine Learning (cs.LG)
Cite as: arXiv:2209.13226 [stat.ML]
  (or arXiv:2209.13226v3 [stat.ML] for this version)

Submission history

From: Shirin Goshtasbpour [view email]
[v1] Tue, 27 Sep 2022 07:58:25 GMT (6971kb,D)
[v2] Thu, 6 Oct 2022 15:23:58 GMT (6971kb,D)
[v3] Sat, 8 Oct 2022 12:01:59 GMT (6971kb,D)

Link back to: arXiv, form interface, contact.