Volume 19, 2007Conference Oxford sur les méthodes de Monte Carlo séquentielles
|Page(s)||6 - 11|
|Published online||30 October 2007|
On the use of sequential Monte Carlo methods for approximating smoothing functionals, with application to fixed parameter estimation
Center for Mathematical Sciences, Lund University, Sweden
2 Laboratoire Traitement et Communication de l'Information, CNRS / GET Télécom Paris, France
3 Centre de Mathématiques APpliquées, Ecole Polytechnique, France
Sequential Monte Carlo (SMC) methods have demonstrated a strong potential for inference on the state variables in Bayesian dynamic models. In this context, it is also often needed to calibrate model parameters. To do so, we consider block maximum likelihood estimation based either on EM (Expectation-Maximization) or gradient methods. In this approach, the key ingredient is the computation of smoothed sum functionals of the hidden states, for a given value of the model parameters. It has been observed by several authors that using standard SMC methods for this smoothing task requires a substantial number of particles and may be unreliable for larger observation sample sizes. We introduce a simple variant of the basic sequential smoothing approach based on forgetting ideas. This modification, which is transparent in terms of computation time, reduces the variability of the approximation of the sum functional. Under suitable regularity assumptions, it is shown that this modification indeed allows a tighter control of the Lp error of the approximation.
Mathematics Subject Classification: 62M20 / 93E25
Key words: Sequential Monte Carlo / Filtering and Smoothing / Parameter Estimation
© EDP Sciences, ESAIM, 2007
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.