Approximating marginals using discrete energy minimization

F. Korc, V. Kolmogorov, C. Lampert, Approximating Marginals Using Discrete Energy Minimization, IST Austria, 2012.

Download
OA 618.74 KB

Technical Report | Published | English
Series Title
IST Austria Technical Report
Abstract
We consider the problem of inference in agraphical model with binary variables. While in theory it is arguably preferable to compute marginal probabilities, in practice researchers often use MAP inference due to the availability of efficient discrete optimization algorithms. We bridge the gap between the two approaches by introducing the Discrete Marginals technique in which approximate marginals are obtained by minimizing an objective function with unary and pair-wise terms over a discretized domain. This allows the use of techniques originally devel-oped for MAP-MRF inference and learning. We explore two ways to set up the objective function - by discretizing the Bethe free energy and by learning it from training data. Experimental results show that for certain types of graphs a learned function can out-perform the Bethe approximation. We also establish a link between the Bethe free energy and submodular functions.
Publishing Year
Date Published
2012-07-23
Page
13
ISSN
IST-REx-ID

Cite this

Korc F, Kolmogorov V, Lampert C. Approximating Marginals Using Discrete Energy Minimization. IST Austria; 2012. doi:10.15479/AT:IST-2012-0003
Korc, F., Kolmogorov, V., & Lampert, C. (2012). Approximating marginals using discrete energy minimization. IST Austria. https://doi.org/10.15479/AT:IST-2012-0003
Korc, Filip, Vladimir Kolmogorov, and Christoph Lampert. Approximating Marginals Using Discrete Energy Minimization. IST Austria, 2012. https://doi.org/10.15479/AT:IST-2012-0003.
F. Korc, V. Kolmogorov, and C. Lampert, Approximating marginals using discrete energy minimization. IST Austria, 2012.
Korc F, Kolmogorov V, Lampert C. 2012. Approximating marginals using discrete energy minimization, IST Austria, 13p.
Korc, Filip, et al. Approximating Marginals Using Discrete Energy Minimization. IST Austria, 2012, doi:10.15479/AT:IST-2012-0003.
Main File(s)
Access Level
OA Open Access
Last Uploaded
2018-12-12T11:53:29Z


Export

Marked Publications

Open Data IST Research Explorer

Search this title in

Google Scholar