Expectation Propagation on the Maximum of Correlated Normal Variables - Statistics > Machine LearningReportar como inadecuado




Expectation Propagation on the Maximum of Correlated Normal Variables - Statistics > Machine Learning - Descarga este documento en PDF. Documentación en PDF para descargar gratis. Disponible también para leer online.

Abstract: Many inference problems involving questions of optimality ask for the maximumor the minimum of a finite set of unknown quantities. This technical reportderives the first two posterior moments of the maximum of two correlatedGaussian variables and the first two posterior moments of the two generatingvariables corresponding to Gaussian approximations minimizing relativeentropy. It is shown how this can be used to build a heuristic approximationto the maximum relationship over a finite set of Gaussian variables, allowingapproximate inference by Expectation Propagation on such quantities.



Autor: Philipp Hennig

Fuente: https://arxiv.org/







Documentos relacionados