The division of graphs, as abstract descriptions of interacting components of a system, is one of the fundamental operations made on graphs, and it can be done using a great variety of algorithms. This work is focused on the division of a graph into two parts using inference methods. One of them is the Belief Propagation algorithm, known for its small runtime, specially when compared with Monte Carlo based methods. However, despite its success on the estimation of the partitioning cost lower bound in most of the situations, when it comes to determine the graph nodes that belong to each group so that such an optimal cost is obtained, the fast Belief Propagation approach tends to fail. In this context, it was developed a partitioning...
L’entropie d’une distribution sur un ensemble de variables aléatoires discrètes est toujours bornée ...
Abstract—The partition function of a factor graph can some-times be accurately estimated by Monte Ca...
Belief propagation over pairwise connected Markov Random Fields has become a widely used approach, a...
In this thesis project we have focused on a kind of graph partitioning problem com- monly called bi-...
Motivated by the belief propagation, we propose a simple anddeterministic message passing algorithm ...
Dans cette thèse, nous étudions le problème de l'inférence bayésienne dans les graphes factoriels, e...
We often encounter probability distributions given as unnormalized products of non-negative function...
Belief propagation on cyclic graphs is an efficient algorithm for computing approximate marginal pro...
Vision tasks, such as segmentation, grouping, recognition, and learning, have a "what-goes-with-what...
The research reported in this thesis focuses on approximation techniques for inference in graphical ...
In this thesis we study two probabilistic models defined on graphs: the Stochastic Block model and t...
Important inference problems in statistical physics, computer vision, error-correcting coding theory...
Belief propagation (BP) on cyclic graphs is an efficient algorithm for computing approximate margina...
In this thesis we propose to use a penalized maximum conditional likelihood criterion to estimate th...
Graph partitioning is the problem of splitting a graph into two or more partitions of fixed sizes wh...
L’entropie d’une distribution sur un ensemble de variables aléatoires discrètes est toujours bornée ...
Abstract—The partition function of a factor graph can some-times be accurately estimated by Monte Ca...
Belief propagation over pairwise connected Markov Random Fields has become a widely used approach, a...
In this thesis project we have focused on a kind of graph partitioning problem com- monly called bi-...
Motivated by the belief propagation, we propose a simple anddeterministic message passing algorithm ...
Dans cette thèse, nous étudions le problème de l'inférence bayésienne dans les graphes factoriels, e...
We often encounter probability distributions given as unnormalized products of non-negative function...
Belief propagation on cyclic graphs is an efficient algorithm for computing approximate marginal pro...
Vision tasks, such as segmentation, grouping, recognition, and learning, have a "what-goes-with-what...
The research reported in this thesis focuses on approximation techniques for inference in graphical ...
In this thesis we study two probabilistic models defined on graphs: the Stochastic Block model and t...
Important inference problems in statistical physics, computer vision, error-correcting coding theory...
Belief propagation (BP) on cyclic graphs is an efficient algorithm for computing approximate margina...
In this thesis we propose to use a penalized maximum conditional likelihood criterion to estimate th...
Graph partitioning is the problem of splitting a graph into two or more partitions of fixed sizes wh...
L’entropie d’une distribution sur un ensemble de variables aléatoires discrètes est toujours bornée ...
Abstract—The partition function of a factor graph can some-times be accurately estimated by Monte Ca...
Belief propagation over pairwise connected Markov Random Fields has become a widely used approach, a...