Separate but equal: Equality in belief propagation for single-cycle graphs

E Cohen, B Rachmut, O Lev, R Zivan - Artificial Intelligence, 2025 - Elsevier
Belief propagation is a widely used, incomplete optimization algorithm whose main
theoretical properties hold only under the assumption that beliefs are not equal …

[HTML][HTML] Governing convergence of Max-sum on DCOPs through damping and splitting

L Cohen, R Galiki, R Zivan - Artificial Intelligence, 2020 - Elsevier
Max-sum is a version of Belief Propagation, used for solving DCOPs. In tree-structured
problems, Max-sum converges to the optimal solution in linear time. Unfortunately, when the …

Taxonomy of dual block-coordinate ascent methods for discrete energy minimization

S Tourani, A Shekhovtsov, C Rother… - International …, 2020 - proceedings.mlr.press
We consider the maximum-a-posteriori inference problem in discrete graphical models and
study solvers based on the dual block-coordinate ascent rule. We map all existing solvers in …

Decomposition bounds for marginal MAP

W Ping, Q Liu, AT Ihler - Advances in neural information …, 2015 - proceedings.neurips.cc
Marginal MAP inference involves making MAP predictions in systems defined with latent
variables or missing information. It is significantly more difficult than pure marginalization …

Beyond trees: Analysis and convergence of belief propagation in graphs with multiple cycles

R Zivan, O Lev, R Galiki - Proceedings of the AAAI Conference on …, 2020 - ojs.aaai.org
Belief propagation, an algorithm for solving problems represented by graphical models, has
long been known to converge to the optimal solution when the graph is a tree. When the …

Bethe learning of graphical models via MAP decoding

K Tang, N Ruozzi, D Belanger… - Artificial Intelligence …, 2016 - proceedings.mlr.press
Many machine learning tasks require fitting probabilistic models over structured objects,
such as pixel grids, matchings, and graph edges. Maximum likelihood estimation (MLE) for …

MS2MP: A min-sum message passing algorithm for motion planning

S Bari, V Gabler, D Wollherr - 2021 IEEE International …, 2021 - ieeexplore.ieee.org
Gaussian Process (GP) formulation of continuous-time trajectory offers a fast solution to the
motion planning problem via probabilistic inference on factor graph. However, often the …

Exactness of approximate MAP inference in continuous MRFs

N Ruozzi - Advances in Neural Information Processing …, 2015 - proceedings.neurips.cc
Computing the MAP assignment in graphical models is generally intractable. As a result, for
discrete graphical models, the MAP problem is often approximated using linear …

Accelerated consensus via min-sum splitting

P Rebeschini, SC Tatikonda - Advances in Neural …, 2017 - proceedings.neurips.cc
Abstract We apply the Min-Sum message-passing protocol to solve the consensus problem
in distributed optimization. We show that while the ordinary Min-Sum algorithm does not …

The effect of asynchronous execution and message latency on max-sum

R Zivan, O Perry, B Rachmut… - … Conference on Principles …, 2021 - drops.dagstuhl.de
Max-sum is a version of belief propagation that was adapted for solving distributed
constraint optimization problems (DCOPs). It has been studied theoretically and empirically …