Dimension reduction in statistical estimation of partially observed multiscale processes

DOI10.1137/16M1085930MaRDI QIDQ4636358zbMATH OpenFDO

Authors Andrew Papanicolaou, Konstantinos Spiliopoulos

Publication date 19 April 2018

Published in SIAM/ASA Journal on Uncertainty Quantification (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1607.06158

zbMATH Keywords

data assimilationdimension reductionfilteringhomogenizationmultiscale diffusionsparameter estimation

Mathematics Subject Classification ID

Non-Markovian processes: hypothesis testing (62M07) Inference from stochastic processes and fuzziness (62M86) Time-scale analysis and singular perturbations in control/observation systems (93C70) Estimation and detection in stochastic control theory (93E10) Filtering in stochastic control theory (93E11)

Abstract: We consider partially observed multiscale diffusion models that are specified up to an unknown vector parameter. We establish for a very general class of test functions that the filter of the original model converges to a filter of reduced dimension. Then, this result is used to justify statistical estimation for the unknown parameters of interest based on the model of reduced dimension but using the original available data. This allows to learn the unknown parameters of interest while working in lower dimensions, as opposed to working with the original high dimensional system. Simulation studies support and illustrate the theoretical results.

Recommendations

• Filtering the Maximum Likelihood for Multiscale Problems
• Dimensional reduction in nonlinear filtering
• Approximate dimension reduction for filtered markov chains
• Particle Filters for Multiscale Diffusions
• Dimensional reduction in nonlinear filtering: a homogenization approach

Cites work

• A PROBLEM IN STOCHASTIC AVERAGING OF NONLINEAR FILTERS
• An applied mathematics perspective on stochastic modelling for climate
• Dimensional reduction in nonlinear filtering
• Dimensional reduction in nonlinear filtering: a homogenization approach
• Efficient nonlinear filtering of a singularly perturbed stochastic hybrid system
• Filtering the Maximum Likelihood for Multiscale Problems
• Fundamentals of stochastic filtering
• scientific article; zbMATH DE number 4211187 (Why is no real title available?)
• scientific article; zbMATH DE number 3688420 (Why is no real title available?)
• scientific article; zbMATH DE number 3627589 (Why is no real title available?)
• scientific article; zbMATH DE number 3274494 (Why is no real title available?)
• Multiscale stochastic volatility for equity, interest rate, and credit derivatives.
• New finite-dimensional filters and smoothers for noisily observed Markov chains
• On Poisson equation and diffusion approximation. II.
• On the Poisson equation and diffusion approximation. I
• Particle Filters for Multiscale Diffusions
• Statistical inference for ergodic diffusion processes.
• Variance Reduction for Particle Filters of Systems With Time Scale Separation
• Weak convergence methods and singularly perturbed stochastic control and filtering problems

Cited in

(7)

• Box dimension estimation of multi-dimensional random fields via wavelet shrinkage
• Effective filtering for multiscale stochastic dynamical systems in Hilbert spaces
• scientific article; zbMATH DE number 1124628 (Why is no real title available?)
• Approximate dimension reduction for filtered markov chains
• Convergence of nonlinear filterings for stochastic dynamical systems with Lévy noises
• Overlapping Clustering Based Technique for Scalable Uncertainty Quantification in Physical Systems
• Convergence of nonlinear filtering for multiscale systems with correlated Lévy noises

This page was built for publication: Dimension reduction in statistical estimation of partially observed multiscale processes

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4636358)