TY - GEN
T1 - Distributed covariance estimation in Gaussian graphical models
AU - Wiesel, Ami
AU - Hero, Alfred O.
PY - 2010
Y1 - 2010
N2 - We consider distributed covariance estimation in Gaussian graphical models. A typical motivation is learning the potential functions for inference via belief propagation in large scale networks. The classical approach based on a centralized maximum likelihood principle is infeasible, and suboptimal distributed alternatives which tradeoff performance with communication costs are required. We begin with a natural solution where each node performs independent estimation of its local covariance with its neighbors. We show that these local solutions are consistent, and can be interpreted as a pseudo-likelihood method. Based on this interpretation, we propose to enhance the performance by introducing additional symmetry constraints. We enforce these using the methodology of the Alternating Direction Method of Multipliers. This results in a flexible message passing protocol between neighboring nodes which can be implemented in large scale networks.
AB - We consider distributed covariance estimation in Gaussian graphical models. A typical motivation is learning the potential functions for inference via belief propagation in large scale networks. The classical approach based on a centralized maximum likelihood principle is infeasible, and suboptimal distributed alternatives which tradeoff performance with communication costs are required. We begin with a natural solution where each node performs independent estimation of its local covariance with its neighbors. We show that these local solutions are consistent, and can be interpreted as a pseudo-likelihood method. Based on this interpretation, we propose to enhance the performance by introducing additional symmetry constraints. We enforce these using the methodology of the Alternating Direction Method of Multipliers. This results in a flexible message passing protocol between neighboring nodes which can be implemented in large scale networks.
UR - http://www.scopus.com/inward/record.url?scp=78650129752&partnerID=8YFLogxK
U2 - 10.1109/SAM.2010.5606735
DO - 10.1109/SAM.2010.5606735
M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???
AN - SCOPUS:78650129752
SN - 9781424489770
T3 - 2010 IEEE Sensor Array and Multichannel Signal Processing Workshop, SAM 2010
SP - 193
EP - 196
BT - 2010 IEEE Sensor Array and Multichannel Signal Processing Workshop, SAM 2010
T2 - 2010 IEEE Sensor Array and Multichannel Signal Processing Workshop, SAM 2010
Y2 - 4 October 2010 through 7 October 2010
ER -