Fokker-Planck particle systems for Bayesian inference: computational approaches

Bayesian inference can be embedded into an appropriately defined dynamics in the space of probability measures. In this paper, we take Brownian motion and its associated Fokker--Planck equation as a starting point for such embeddings and explore several interacting particle approximations. More spec...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Reich, Sebastian (VerfasserIn) , Weissmann, Simon (VerfasserIn)
Dokumenttyp: Article (Journal)
Sprache:Englisch
Veröffentlicht: April 26, 2021
In: SIAM ASA journal on uncertainty quantification
Year: 2021, Jahrgang: 9, Heft: 2, Pages: 446-482
ISSN:2166-2525
DOI:10.1137/19M1303162
Online-Zugang:Verlag, lizenzpflichtig, Volltext: https://doi.org/10.1137/19M1303162
Verlag, lizenzpflichtig, Volltext: https://epubs.siam.org/doi/10.1137/19M1303162
Volltext
Verfasserangaben:Sebastian Reich and Simon Weissmann

MARC

LEADER 00000caa a2200000 c 4500
001 1769435301
003 DE-627
005 20220820041701.0
007 cr uuu---uuuuu
008 210904s2021 xx |||||o 00| ||eng c
024 7 |a 10.1137/19M1303162  |2 doi 
035 |a (DE-627)1769435301 
035 |a (DE-599)KXP1769435301 
035 |a (OCoLC)1341420691 
040 |a DE-627  |b ger  |c DE-627  |e rda 
041 |a eng 
084 |a 27  |2 sdnb 
100 1 |a Reich, Sebastian  |d 1963-  |e VerfasserIn  |0 (DE-588)1078254028  |0 (DE-627)838201679  |0 (DE-576)450178617  |4 aut 
245 1 0 |a Fokker-Planck particle systems for Bayesian inference  |b computational approaches  |c Sebastian Reich and Simon Weissmann 
264 1 |c  April 26, 2021 
300 |a 37 
336 |a Text  |b txt  |2 rdacontent 
337 |a Computermedien  |b c  |2 rdamedia 
338 |a Online-Ressource  |b cr  |2 rdacarrier 
500 |a Gesehen am 04.09.2021 
520 |a Bayesian inference can be embedded into an appropriately defined dynamics in the space of probability measures. In this paper, we take Brownian motion and its associated Fokker--Planck equation as a starting point for such embeddings and explore several interacting particle approximations. More specifically, we consider both deterministic and stochastic interacting particle systems and combine them with the idea of preconditioning by the empirical covariance matrix. In addition to leading to affine invariant formulations which asymptotically speed up convergence, preconditioning allows for gradient-free implementations in the spirit of the ensemble Kalman filter. While such gradient-free implementations have been demonstrated to work well for posterior measures that are nearly Gaussian, we extend their scope of applicability to multimodal measures by introducing localized gradient-free approximations. Numerical results demonstrate the effectiveness of the considered methodologies. 
650 4 |a 62F15 
650 4 |a 65C30 
650 4 |a 65N21 
650 4 |a 65N75 
650 4 |a 90C56 
650 4 |a affine invariance 
650 4 |a Bayesian inverse problems 
650 4 |a Fokker;Planck equation 
650 4 |a gradient flow 
650 4 |a gradient-free sampling methods 
650 4 |a localization 
700 1 |a Weissmann, Simon  |d 1992-  |e VerfasserIn  |0 (DE-588)1225578299  |0 (DE-627)1745296557  |4 aut 
773 0 8 |i Enthalten in  |a Society for Industrial and Applied Mathematics  |t SIAM ASA journal on uncertainty quantification  |d Philadelphia, Pa. : SIAM, 2013  |g 9(2021), 2, Seite 446-482  |h Online-Ressource  |w (DE-627)729754006  |w (DE-600)2690603-X  |w (DE-576)391400703  |x 2166-2525  |7 nnas 
773 1 8 |g volume:9  |g year:2021  |g number:2  |g pages:446-482  |g extent:37  |a Fokker-Planck particle systems for Bayesian inference computational approaches 
856 4 0 |u https://doi.org/10.1137/19M1303162  |x Verlag  |x Resolving-System  |z lizenzpflichtig  |3 Volltext 
856 4 0 |u https://epubs.siam.org/doi/10.1137/19M1303162  |x Verlag  |z lizenzpflichtig  |3 Volltext 
951 |a AR 
992 |a 20210904 
993 |a Article 
994 |a 2021 
998 |g 1225578299  |a Weissmann, Simon  |m 1225578299:Weissmann, Simon  |d 700000  |d 708000  |e 700000PW1225578299  |e 708000PW1225578299  |k 0/700000/  |k 1/700000/708000/  |p 2  |y j 
999 |a KXP-PPN1769435301  |e 3974118845 
BIB |a Y 
SER |a journal 
JSO |a {"title":[{"subtitle":"computational approaches","title":"Fokker-Planck particle systems for Bayesian inference","title_sort":"Fokker-Planck particle systems for Bayesian inference"}],"person":[{"family":"Reich","role":"aut","display":"Reich, Sebastian","given":"Sebastian"},{"family":"Weissmann","role":"aut","display":"Weissmann, Simon","given":"Simon"}],"physDesc":[{"extent":"37 S."}],"id":{"eki":["1769435301"],"doi":["10.1137/19M1303162"]},"name":{"displayForm":["Sebastian Reich and Simon Weissmann"]},"language":["eng"],"type":{"bibl":"article-journal","media":"Online-Ressource"},"recId":"1769435301","relHost":[{"recId":"729754006","pubHistory":["1.2013 -"],"note":["Gesehen am 21.05.13"],"origin":[{"publisher":"SIAM","dateIssuedDisp":"2013-","dateIssuedKey":"2013","publisherPlace":"Philadelphia, Pa."}],"id":{"issn":["2166-2525"],"zdb":["2690603-X"],"eki":["729754006"]},"name":{"displayForm":["Society for Industrial and Applied Mathematics ; American Statistical Association"]},"language":["eng"],"titleAlt":[{"title":"JUQ"},{"title":"Journal on uncertainty quantification"}],"corporate":[{"display":"Society for Industrial and Applied Mathematics","role":"aut"},{"display":"American Statistical Association","role":"isb"}],"physDesc":[{"extent":"Online-Ressource"}],"disp":"Society for Industrial and Applied MathematicsSIAM ASA journal on uncertainty quantification","title":[{"title":"SIAM ASA journal on uncertainty quantification","title_sort":"SIAM ASA journal on uncertainty quantification"}],"part":{"text":"9(2021), 2, Seite 446-482","pages":"446-482","year":"2021","issue":"2","extent":"37","volume":"9"},"type":{"bibl":"periodical","media":"Online-Ressource"}}],"note":["Gesehen am 04.09.2021"],"origin":[{"dateIssuedKey":"2021","dateIssuedDisp":" April 26, 2021"}]} 
SRT |a REICHSEBASFOKKERPLAN2620