[PDF][PDF] High-dimensional Gaussian graphical model selection: Walk summability and local separation criterion
We consider the problem of high-dimensional Gaussian graphical model selection. We
identify a set of graphs for which an efficient estimation algorithm exists, and this algorithm is …
identify a set of graphs for which an efficient estimation algorithm exists, and this algorithm is …
The Bethe permanent of a nonnegative matrix
PO Vontobel - IEEE Transactions on Information Theory, 2012 - ieeexplore.ieee.org
It has recently been observed that the permanent of a nonnegative square matrix, ie, of a
square matrix containing only nonnegative real entries, can very well be approximated by …
square matrix containing only nonnegative real entries, can very well be approximated by …
Message-Passing Algorithms for Inference and Optimization: “Belief Propagation” and “Divide and Concur”
JS Yedidia - Journal of Statistical Physics, 2011 - Springer
Message-passing algorithms can solve a wide variety of optimization, inference, and
constraint satisfaction problems. The algorithms operate on factor graphs that visually …
constraint satisfaction problems. The algorithms operate on factor graphs that visually …
Divide and concur and difference-map BP decoders for LDPC codes
JS Yedidia, Y Wang, SC Draper - IEEE Transactions on …, 2011 - ieeexplore.ieee.org
The “Divide and Concur”(DC) algorithm introduced by Gravel and Elser can be considered a
competitor to the belief propagation (BP) algorithm, in that both algorithms can be applied to …
competitor to the belief propagation (BP) algorithm, in that both algorithms can be applied to …
Tightening fractional covering upper bounds on the partition function for high-order region graphs
In this paper we present a new approach for tightening upper bounds on the partition
function. Our upper bounds are based on fractional covering bounds on the entropy function …
function. Our upper bounds are based on fractional covering bounds on the entropy function …
Message error analysis of loopy belief propagation
X Shi, D Schonfeld, D Tuninetti - 2010 IEEE International …, 2010 - ieeexplore.ieee.org
The loopy belief propagation algorithm (LBP) is known to perform extremely well in many
practical problems of probability inference and learning on graphical models, even in …
practical problems of probability inference and learning on graphical models, even in …
[PDF][PDF] High-dimensional Gaussian graphical model selection: Tractable graph families
We consider the problem of high-dimensional Gaussian graphical model selection. We
identify a set of graphs for which an efficient estimation algorithm exists, and this algorithm is …
identify a set of graphs for which an efficient estimation algorithm exists, and this algorithm is …
Message-passing in stochastic processing networks
D Shah - Surveys in Operations Research and Management …, 2011 - Elsevier
Simple, distributed and iterative algorithms, popularly known as message-passing, have
become the architecture of choice for emerging infrastructure networks and the canonical …
become the architecture of choice for emerging infrastructure networks and the canonical …
A combinatorial characterization of the Bethe and the Kikuchi partition functions
PO Vontobel - 2011 Information Theory and Applications …, 2011 - ieeexplore.ieee.org
The partition function contains valuable information about a graphical model. Unfortunately,
the exact partition function is very often intractable, and so suitable approximations are …
the exact partition function is very often intractable, and so suitable approximations are …
Unconstrained minimization of quadratic functions via min-sum
N Ruozzi, S Tatikonda - 2010 44th Annual Conference on …, 2010 - ieeexplore.ieee.org
Gaussian belief propagation is an iterative algorithm for computing the mean of a
multivariate Gaussian distribution. Equivalently, the min-sum algorithm can be used to …
multivariate Gaussian distribution. Equivalently, the min-sum algorithm can be used to …