Revisiting the Limits of MAP Inference by MWSS on Perfect Graphs

Adrian Weller
Proceedings of the Eighteenth International Conference on Artificial Intelligence and Statistics, PMLR 38:1061-1069, 2015.

Abstract

A recent, promising approach to identifying a configuration of a discrete graphical model with highest probability (termed MAP inference) is to reduce the problem to finding a maximum weight stable set (MWSS) in a derived weighted graph, which, if perfect, allows a solution to be found in polynomial time. Weller and Jebara (2013) investigated the class of binary pairwise models where this method may be applied. However, their analysis made a seemingly innocuous assumption which simplifies analysis but led to only a subset of possible reparameterizations being considered. Here we introduce novel techniques and consider all cases, demonstrating that this greatly expands the set of tractable models. We provide a simple, exact characterization of the new, enlarged set and show how such models may be efficiently identified, thus settling the power of the approach on this class.

Cite this Paper


BibTeX
@InProceedings{pmlr-v38-weller15, title = {{Revisiting the Limits of MAP Inference by MWSS on Perfect Graphs}}, author = {Weller, Adrian}, booktitle = {Proceedings of the Eighteenth International Conference on Artificial Intelligence and Statistics}, pages = {1061--1069}, year = {2015}, editor = {Lebanon, Guy and Vishwanathan, S. V. N.}, volume = {38}, series = {Proceedings of Machine Learning Research}, address = {San Diego, California, USA}, month = {09--12 May}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v38/weller15.pdf}, url = {https://proceedings.mlr.press/v38/weller15.html}, abstract = {A recent, promising approach to identifying a configuration of a discrete graphical model with highest probability (termed MAP inference) is to reduce the problem to finding a maximum weight stable set (MWSS) in a derived weighted graph, which, if perfect, allows a solution to be found in polynomial time. Weller and Jebara (2013) investigated the class of binary pairwise models where this method may be applied. However, their analysis made a seemingly innocuous assumption which simplifies analysis but led to only a subset of possible reparameterizations being considered. Here we introduce novel techniques and consider all cases, demonstrating that this greatly expands the set of tractable models. We provide a simple, exact characterization of the new, enlarged set and show how such models may be efficiently identified, thus settling the power of the approach on this class.} }
Endnote
%0 Conference Paper %T Revisiting the Limits of MAP Inference by MWSS on Perfect Graphs %A Adrian Weller %B Proceedings of the Eighteenth International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2015 %E Guy Lebanon %E S. V. N. Vishwanathan %F pmlr-v38-weller15 %I PMLR %P 1061--1069 %U https://proceedings.mlr.press/v38/weller15.html %V 38 %X A recent, promising approach to identifying a configuration of a discrete graphical model with highest probability (termed MAP inference) is to reduce the problem to finding a maximum weight stable set (MWSS) in a derived weighted graph, which, if perfect, allows a solution to be found in polynomial time. Weller and Jebara (2013) investigated the class of binary pairwise models where this method may be applied. However, their analysis made a seemingly innocuous assumption which simplifies analysis but led to only a subset of possible reparameterizations being considered. Here we introduce novel techniques and consider all cases, demonstrating that this greatly expands the set of tractable models. We provide a simple, exact characterization of the new, enlarged set and show how such models may be efficiently identified, thus settling the power of the approach on this class.
RIS
TY - CPAPER TI - Revisiting the Limits of MAP Inference by MWSS on Perfect Graphs AU - Adrian Weller BT - Proceedings of the Eighteenth International Conference on Artificial Intelligence and Statistics DA - 2015/02/21 ED - Guy Lebanon ED - S. V. N. Vishwanathan ID - pmlr-v38-weller15 PB - PMLR DP - Proceedings of Machine Learning Research VL - 38 SP - 1061 EP - 1069 L1 - http://proceedings.mlr.press/v38/weller15.pdf UR - https://proceedings.mlr.press/v38/weller15.html AB - A recent, promising approach to identifying a configuration of a discrete graphical model with highest probability (termed MAP inference) is to reduce the problem to finding a maximum weight stable set (MWSS) in a derived weighted graph, which, if perfect, allows a solution to be found in polynomial time. Weller and Jebara (2013) investigated the class of binary pairwise models where this method may be applied. However, their analysis made a seemingly innocuous assumption which simplifies analysis but led to only a subset of possible reparameterizations being considered. Here we introduce novel techniques and consider all cases, demonstrating that this greatly expands the set of tractable models. We provide a simple, exact characterization of the new, enlarged set and show how such models may be efficiently identified, thus settling the power of the approach on this class. ER -
APA
Weller, A.. (2015). Revisiting the Limits of MAP Inference by MWSS on Perfect Graphs. Proceedings of the Eighteenth International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 38:1061-1069 Available from https://proceedings.mlr.press/v38/weller15.html.

Related Material