Original title: Score matching filters for Gaussian Markov random fields with a linear model of the precision matrix
Authors: Turčičová, Marie ; Mandel, J. ; Eben, Kryštof
Document type: Research reports
Year: 2021
Language: eng
Series: Technical Report, volume: V-1284
Abstract: We present an ensemble filter that provides a rigorous covariance regularization when the underlying random field is Gaussian Markov. We use a linear model for the precision matrix (inverse of covariance) and estimate its parameters together with the analysis mean by the Score Matching method. This procedure provides an explicit expression for parameter estimators. The resulting analysis step formula is the same as in the traditional ensemble Kalman filter.
Keywords: covariance modelling; ensemble filter; Gaussian Markov random field; Score matching
Project no.: TL01000238
Funding provider: GA TA ČR
Note: Související webová stránka: https://www.aimsciences.org/article/doi/10.3934/fods.2021030

Institution: Institute of Computer Science AS ČR (web)
Document availability information: Fulltext is available at the institute of the Academy of Sciences.
Original record: http://hdl.handle.net/11104/0319459

Permalink: http://www.nusl.cz/ntk/nusl-438727


The record appears in these collections:
Research > Institutes ASCR > Institute of Computer Science
Reports > Research reports