Bayesian Filtering with Intractable Likelihood using Sequential MCMC

Abstract : We develop a sequential estimation methodology for a class of non- linear, non-Gaussian state space models in which the observation process is intractable to express in closed form, but trivial to simu- late. In addition we consider models in which the latent state vector and the observation vector are very high dimensional. To overcome these two difficulties we propose the class of Sequential Markov chain Monte Carlo (SMCMC) algorithms in which we incorporate a component of Approximate Bayesian Computation (ABC). In do- ing so we tackle both the curse of dimensionality via the SMCMC and the intractability of the likelihood via the ABC component. We demonstrate how the proposed algorithm outperforms alternative ap- proaches in two challenging state space model examples.
Type de document :
Communication dans un congrès
IEEE International Conference on Acoustics, Speech, and Signal Processing, May 2013, Vancouver, Canada. pp.1-5, 2013
Liste complète des métadonnées

https://hal-imt.archives-ouvertes.fr/hal-00813176
Contributeur : François Septier <>
Soumis le : lundi 15 avril 2013 - 11:17:26
Dernière modification le : jeudi 12 avril 2018 - 16:25:49

Identifiants

  • HAL Id : hal-00813176, version 1

Collections

Citation

François Septier, Gareth W. Peters, Ido Nevat. Bayesian Filtering with Intractable Likelihood using Sequential MCMC. IEEE International Conference on Acoustics, Speech, and Signal Processing, May 2013, Vancouver, Canada. pp.1-5, 2013. 〈hal-00813176〉

Partager

Métriques

Consultations de la notice

210