DocumentCode
2117954
Title
Bayesian Generalized Linear Models in a Terabyte World
Author
Zoeter, Onno
Author_Institution
Microsoft Res. Cambridge, Cambridge
fYear
2007
fDate
27-29 Sept. 2007
Firstpage
435
Lastpage
440
Abstract
This paper introduces extremely fast approximate inference schemes for Bayesian treatments of dynamic generalized linear models. The approximations are tailored variants of quadrature EP. The first forward pass of this fixed point iteration algorithm can be interpreted as a one-step unscented Kalman filter. For on-line applications this filter can handle tens of thousands of updates a second on a current day desktop machine.
Keywords
Bayes methods; inference mechanisms; regression analysis; Bayesian generalized linear models; Bayesian treatments; approximate inference schemes; desktop machine; fixed point iteration algorithm; one-step unscented Kalman filter; Bayesian methods; Filters; Gaussian distribution; Inference algorithms; Kernel; Large-scale systems; Linear regression; Sampling methods; Statistics; Vectors;
fLanguage
English
Publisher
ieee
Conference_Titel
Image and Signal Processing and Analysis, 2007. ISPA 2007. 5th International Symposium on
Conference_Location
Istanbul
ISSN
1845-5921
Print_ISBN
978-953-184-116-0
Type
conf
DOI
10.1109/ISPA.2007.4383733
Filename
4383733
Link To Document