Title :
GMRF Variance Approximation using Splicedwavelet Bases
Author :
Malioutov, Dmitry M. ; Johnson, J.K. ; Willsky, Alan S.
Author_Institution :
Lab. for Inf. & Decision Syst., Massachusetts Inst. of Technol., Cambridge, MA, USA
Abstract :
We consider the problem of computing variances in large-scale Gauss-Markov random field (GMRF) models. In our prior work we considered the short-range correlation case, and we proposed a simple low-rank method which computes approximate variances with linear complexity in the number of nodes. In addition to its low complexity, the method has good guarantees on the quality of the approximation. In this paper we extend our method and analysis using a wavelet-based multi-scale approach which is applicable to models with much longer correlation lengths.
Keywords :
Gaussian processes; Markov processes; wavelet transforms; GMRF variance approximation; large-scale Gauss-Markov random field models; linear complexity; short-range correlation case; spliced wavelet bases; wavelet-based multi-scale approach; Application software; Discrete wavelet transforms; Gaussian approximation; Laboratories; Large-scale systems; Linear approximation; Military computing; Random variables; Sparse matrices; Symmetric matrices; GMRF; approximate variances; wavelets;
Conference_Titel :
Acoustics, Speech and Signal Processing, 2007. ICASSP 2007. IEEE International Conference on
Conference_Location :
Honolulu, HI
Print_ISBN :
1-4244-0727-3
DOI :
10.1109/ICASSP.2007.366876