Score matching filters for Gaussian Markov random fields with a linear model of the precision matrix

0541945 - ÚI 2022 CZ eng V - Research Report
Turčičová, Marie - Mandel, J. - Eben, Kryštof
Score matching filters for Gaussian Markov random fields with a linear model of the precision matrix.
Prague: ICS CAS, 2021. 30 s. Technical Report, V-1284.
R&D Projects: GA TA ČR(CZ) TL01000238
Institutional support: RVO:67985807
Keywords : Score matching * ensemble filter * Gaussian Markov random field * covariance modelling
https://www.aimsciences.org/article/doi/10.3934/fods.2021030

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.
Permanent Link: http://hdl.handle.net/11104/0319459