• Aucun résultat trouvé

Un cadre pour les méthodes de Monte Carlo adaptatives

N/A
N/A
Protected

Academic year: 2021

Partager "Un cadre pour les méthodes de Monte Carlo adaptatives"

Copied!
2
0
0

Texte intégral

(1)

HAL Id: inria-00510322

https://hal.inria.fr/inria-00510322

Submitted on 18 Aug 2010

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Un cadre pour les méthodes de Monte Carlo adaptatives

Jérôme Lelong

To cite this version:

Jérôme Lelong. Un cadre pour les méthodes de Monte Carlo adaptatives. Journées MAS et Journée en l’honneur de Jacques Neveu, Aug 2010, Talence, France. �inria-00510322�

(2)

Journées MAS 2010, Bordeaux

Session : Algorithmes Stochastiques

Un cadre pour les méthodes de Monte Carlo adaptatives

par Jérome Lelong

Adaptive Monte Carlo methods are powerful variance reduction techniques.

In this work, we propose a mathematical setting which greatly relaxes the assumptions needed by for the adaptive importance sampling techniques pre- sented by Arouna in 2003. We establish the convergence and asymptotic nor- mality of the adaptive Monte Carlo estimator under local assumptions which are easily veriable in practice. We present one way of approximating the op- timal importance sampling parameter using a randomly truncated stochastic algorithm. Finally, we apply this technique to the valuation of nancial deri- vatives and our numerical experiments show a signicant variance reduction.

Adresse :

Jérome Lelong

ENSIMAG- Laboratoire Jean Kuntzmann, Grenoble E-mail : [email protected]

Session : Algorithmes Stochastiques

Références

Documents relatifs

Variance Reduction for Generalized Likelihood Ratio Method By Conditional Monte Carlo and Randomized Quasi-Monte Carlo... Variance Reduction for Generalized Likelihood Ratio Method

Although all the adap- tive algorithms satisfy the same central limit theorem, they may behave very differently in practice, in particular adaptive algorithms using a

To tackle large dimensions, other methods have been devised that are based on a non-independent sampling: Markov chain Monte Carlo methods (MCMC).. The prototype of MCMC methods is

Uniform sampling Update error and pixel probabilities Adaptive sampling Uniform sampling. Comparison with Tsallis entropy [Xu, Sbert, Xinh and

FLØTTUM,,,, les sources d’inspiration de cette recherche, pour leur générosité et pour le rôle qu’elles ont joué dans l’élaboration de ce travail. Un grand merci à ma

A powerful Monte Carlo variance reduction technique introduced in (Cao, M. Hussaini, and Zhang, 2004) uses local derivatives to accelerate Monte Carlo

It is of course possible to generalize the little example from a probability λ = 1/6 to a general value of λ, and from a single spin to a general statistical mechanics problem

This approach can be generalized to obtain a continuous CDF estimator and then an unbiased density estimator, via the likelihood ratio (LR) simulation-based derivative estimation