Skip to Main content Skip to Navigation
New interface
Reports (Research report)

Pairwise MRF Calibration by Perturbation of the Bethe Reference Point

Cyril Furtlehner 1 Yufei Han 2 Jean-Marc Lasgouttes 2 Victorin Martin 2 
1 TAO - Machine Learning and Optimisation
LRI - Laboratoire de Recherche en Informatique, UP11 - Université Paris-Sud - Paris 11, Inria Saclay - Ile de France, CNRS - Centre National de la Recherche Scientifique : UMR8623
Abstract : We investigate different ways of generating approximate solutions to the inverse problem of pairwise Markov random field (MRF) model learning. We focus mainly on the inverse Ising problem, but discuss also the somewhat related inverse Gaussian problem. In both cases, the belief propagation algorithm can be used in closed form to perform inference tasks. Our approach consists in taking the Bethe mean-field solution as a reference point for further perturbation procedures. We remark indeed that both the natural gradient and the best link to be added to a maximum spanning tree (MST) solution can be computed analytically. These observations open the way to many possible algorithms, able to find approximate sparse solutions compatible with belief propagation inference procedures. Experimental tests on various datasets with refined $L_0$ or $L_1$ regularization procedures indicate that this approach may be a competitive and useful alternative to existing ones.
Complete list of metadata

Cited literature [36 references]  Display  Hide  Download
Contributor : Cyril Furtlehner Connect in order to contact the contributor
Submitted on : Thursday, October 18, 2012 - 5:02:37 PM
Last modification on : Friday, October 28, 2022 - 3:28:46 AM
Long-term archiving on: : Saturday, January 19, 2013 - 3:40:51 AM


Files produced by the author(s)


  • HAL Id : hal-00743334, version 1


Cyril Furtlehner, Yufei Han, Jean-Marc Lasgouttes, Victorin Martin. Pairwise MRF Calibration by Perturbation of the Bethe Reference Point. [Research Report] RR-8059, INRIA. 2012, pp.35. ⟨hal-00743334⟩



Record views


Files downloads