Multilevel Monte Carlo for Smoothing via Transport Methods

DOI10.1137/17M1156071zbMATH Open1395.62256arXiv1711.02836OpenAlexW2963868359WikidataQ129412110 ScholiaQ129412110MaRDI QIDQ4580285FDOQ4580285

Jeremie Houssineau, Sumeetpal S. Singh, Ajay Jasra

Publication date: 14 August 2018

Published in: SIAM Journal on Scientific Computing (Search for Journal in Brave)

Abstract: In this article we consider recursive approximations of the smoothing distribution associated to partially observed stochastic differential equations (SDEs), which are observed discretely in time. Such models appear in a wide variety of applications including econometrics, finance and engineering. This problem is notoriously challenging, as the smoother is not available analytically and hence require numerical approximation. This usually consists by applying a time-discretization to the SDE, for instance the Euler method, and then applying a numerical (e.g. Monte Carlo) method to approximate the smoother. This has lead to a vast literature on methodology for solving such problems, perhaps the most popular of which is based upon the particle filter (PF) e.g. [9]. In the context of filtering for this class of problems, it is well-known that the particle filter can be improved upon in terms of cost to achieve a given mean squared error (MSE) for estimates. This in the sense that the computational effort can be reduced to achieve this target MSE, by using multilevel (ML) methods [12, 13, 18], via the multilevel particle filter (MLPF) [16, 20, 21]. For instance, to obtain a MSE of $mathcalO(epsilo{n}^2)$ for some $epsilon>0$ when approximating filtering distributions associated with Euler-discretized diffusions with constant diffusion coefficients, the cost of the PF is $mathcalO(epsilo{n}^{-3})$ while the cost of the MLPF is $mathcalO(epsilo{n}^{-2}log(epsilon{)}^2)$. In this article we consider a new approach to replace the particle filter, using transport methods in [27]. In the context of filtering, one expects that the proposed method improves upon the MLPF by yielding, under assumptions, a MSE of $mathcalO(epsilo{n}^2)$ for a cost of $mathcalO(epsilo{n}^{-2})$. This is established theoretically in an "ideal" example and numerically in numerous examples.

Full work available at URL: https://arxiv.org/abs/1711.02836

zbMATH Keywords

stochastic differential equationmultilevel Monte Carlotransport map

Mathematics Subject Classification ID

Markov processes: estimation; hidden Markov models (62M05) Diffusion processes (60J60)

Cites Work

• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Inference in hidden Markov models.
• Optimal Transport
• Exact and Computationally Efficient Likelihood-Based Estimation for Discretely Observed Diffusion Processes (with Discussion)
• On particle methods for parameter estimation in state-space models
• Multilevel Monte Carlo Path Simulation
• Closed-form likelihood expansions for multivariate diffusions
• Exact simulation of diffusions
• Antithetic multilevel Monte Carlo estimation for multi-dimensional SDEs without Lévy area simulation
• Multi-index Monte Carlo: when sparsity meets sampling
• Bayesian inference with optimal maps
• Fundamentals of stochastic filtering
• Analyzing multi-level Monte Carlo for options with non-globally Lipschitz payoff
• Particle Filters for Partially Observed Diffusions
• Multilevel Monte Carlo Methods
• Multilevel Particle Filters
• Multilevel particle filters: normalizing constant estimation
• Multilevel ensemble transform particle filtering
• On coupling particle filter trajectories
• A Multiscale Strategy for Bayesian Inference Using Transport Maps
• Unbiased multi-index Monte Carlo

Cited In (4)

• A Wasserstein coupled particle filter for multilevel estimation
• Multilevel particle filters for Lévy-driven stochastic differential equations
• Coupling Techniques for Nonlinear Ensemble Filtering
• Advanced Multilevel Monte Carlo Methods

This page was built for publication: Multilevel Monte Carlo for Smoothing via Transport Methods

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4580285)