English
An introduction to particle filters
Multi angle
Auteurs : Chopin, Nicolas (Auteur de la Conférence)
CIRM (Editeur )
60J22 - Computational methods in Markov chains
62D05 - Sampling theory, sample surveys
62F15 - Bayesian inference
62M05 - Markov processes: estimation
62M20 - Prediction; filtering (statistics)
65C05 - Monte Carlo methods
CIRM (Editeur )
Loading the player...
|
Résumé :
This course will give a gentle introduction to SMC (Sequential Monte Carlo algorithms):
• motivation: state-space (hidden Markov) models, sequential analysis of such models; non-sequential problems that may be tackled using SMC.
• Formalism: Markov kernels, Feynman-Kac distributions.
• Monte Carlo tricks: importance sampling and resampling
• standard particle filters: bootstrap, guided, auxiliary
• maximum likelihood estimation of state-stace models
• Bayesian estimation of these models: PMCMC, SMC$^2$.
Keywords : sequential Monte Carlo methods; state-space model; resample-move algorithm; particle filtering; maximum likelihood estimators (MLE); Bayesian estimation
Codes MSC : • motivation: state-space (hidden Markov) models, sequential analysis of such models; non-sequential problems that may be tackled using SMC.
• Formalism: Markov kernels, Feynman-Kac distributions.
• Monte Carlo tricks: importance sampling and resampling
• standard particle filters: bootstrap, guided, auxiliary
• maximum likelihood estimation of state-stace models
• Bayesian estimation of these models: PMCMC, SMC$^2$.
Keywords : sequential Monte Carlo methods; state-space model; resample-move algorithm; particle filtering; maximum likelihood estimators (MLE); Bayesian estimation
60J22 - Computational methods in Markov chains
62D05 - Sampling theory, sample surveys
62F15 - Bayesian inference
62M05 - Markov processes: estimation
62M20 - Prediction; filtering (statistics)
65C05 - Monte Carlo methods
|
Informations sur la Vidéo
Réalisateur : Hennenfent, GuillaumeLangue : Anglais Date de publication : 12/11/2018 Date de captation : 25/10/2018 Sous collection : Research School arXiv category : Computation ; Statistics Domaine : Probability & Statistics Format : MP4 (.mp4) - HD Durée : 01:59:38 Audience : Researchers ; Graduate Students Download : https://videos.cirm-math.fr/2018-10-25_Chopin.mp4 
|
Informations sur la Rencontre
Nom de la rencontre : Jean-Morlet chair: Masterclass in Bayesian statistics / Chaire Jean-Morlet : École de statistique bayésienneOrganisateurs de la rencontre : Chopin, Nicolas ; Mengersen, Kerrie ; Pommeret, Denys ; Pudlo, Pierre ; Robert, Christian P. ; Ryder, Robin Dates : 22/10/2018 - 26/10/2018 Année de la rencontre : 2018 URL Congrès : https://www.chairejeanmorlet.com/1854.html
Données de citation
DOI : 10.24350/CIRM.V.19468603Citer cette vidéo: Chopin, Nicolas (2018). An introduction to particle filters. CIRM. Audiovisual resource. doi:10.24350/CIRM.V.19468603 URI : http://dx.doi.org/10.24350/CIRM.V.19468603 |
Voir aussi
- [Multi angle] Model assessment, selection and averaging / Auteur de la Conférence Vehtari, Aki.
- [Multi angle] Bayesian computation with INLA / Auteur de la Conférence Rue, Havard.
- [Multi angle] Bayesian computational methods / Auteur de la Conférence Robert, Christian P..
- [ Post-edited] Bayesian modelling / Auteur de la Conférence Mengersen, Kerrie.
Bibliographie
