Back
Fast Convergent Algorithms for Expectation Propagation Approximate Bayesian Inference
We propose a novel algorithm to solve the expectation propagation relaxation of Bayesian inference for continuous-variable graphical models. In contrast to most previous algorithms, our method is provably convergent. By marrying convergent EP ideas from (Opper&Winther 05) with covariance decoupling techniques (Wipf&Nagarajan 08, Nickisch&Seeger 09), it runs at least an order of magnitude faster than the most commonly used EP solver.
@techreport{6995, title = {Fast Convergent Algorithms for Expectation Propagation Approximate Bayesian Inference}, abstract = {We propose a novel algorithm to solve the expectation propagation relaxation of Bayesian inference for continuous-variable graphical models. In contrast to most previous algorithms, our method is provably convergent. By marrying convergent EP ideas from (Opper&Winther 05) with covariance decoupling techniques (Wipf&Nagarajan 08, Nickisch&Seeger 09), it runs at least an order of magnitude faster than the most commonly used EP solver.}, organization = {Max-Planck-Gesellschaft}, institution = {Max Planck Institute for Biological Cybernetics}, school = {Biologische Kybernetik}, month = dec, year = {2010}, slug = {6995}, author = {Seeger, M. and Nickisch, H.}, month_numeric = {12} }