Rough McKean-Vlasov dynamics for robust ensemble Kalman filtering
arXiv:2107.06621 · doi:10.1214/23-AAP1957
Abstract
Motivated by the challenge of incorporating data into misspecified and multiscale dynamical models, we study a McKean-Vlasov equation that contains the data stream as a common driving rough path. This setting allows us to prove well-posedness as well as continuity with respect to the driver in an appropriate rough-path topology. The latter property is key in our subsequent development of a robust data assimilation methodology: We establish propagation of chaos for the associated interacting particle system, which in turn is suggestive of a numerical scheme that can be viewed as an extension of the ensemble Kalman filter to a rough-path framework. Finally, we discuss a data-driven method based on subsampling to construct suitable rough path lifts and demonstrate the robustness of our scheme in a number of numerical experiments related to parameter estimation problems in multiscale contexts.
arXiv comment44 pages, 7 figures
References in corpus (34)
- Parameter Estimation for Multiscale Diffusions
- Supervised learning from noisy observations: Combining machine-learning techniques with data assimilation
- Rough paths, Signatures and the modelling of functions on streams
- Propagation of chaos for interacting particles subject to environmental noise
- Learning from the past, predicting the statistics for the future, learning an evolving system
- A Milstein-type scheme without Levy area terms for SDEs driven by fractional Brownian motion
- Long-time stability and accuracy of the ensemble Kalman-Bucy filter for fully observed processes and small measurement noise
- Ensemble Kalman Inversion: mean-field limit and convergence analysis
- Data Assimilation: The Schrödinger Perspective
- Discretely sampled signals and the rough Hoff process
- On the geometry of Stein variational gradient descent
- Stochastic nonlinear Fokker-Planck equations
- Robust filtering: Correlated noise and multidimensional observation
- Evolving Communities with Individual Preferences
- Stochastic partial differential equations: a rough path view
- Pathwise McKean-Vlasov Theory with Additive Noise
- Dimensional reduction in nonlinear filtering: A homogenization approach
- McKean-Vlasov SDEs in nonlinear filtering
- Semi-Parametric Drift and Diffusion Estimation for Multiscale Diffusions
- Particle filtering in high-dimensional chaotic systems
- Multiscale systems, homogenization, and rough paths
- Bayesian inference of ocean diffusivity from Lagrangian trajectory data
- On the Mathematical Theory of Ensemble (Linear-Gaussian) Kalman-Bucy Filtering
- On the continuous time limit of the Ensemble Kalman Filter
- State and Parameter Estimation from Observed Signal Increments
- Statistical properties for flows with unbounded roof function, including the Lorenz attractor
- Analysis of a localised nonlinear Ensemble Kalman Bucy Filter with complete and accurate observations
- Mean field limit of Ensemble Square Root Filters -- discrete and continuous time
- Ensemble Kalman Inversion for nonlinear problems: weights, consistency, and variance bounds
- Rough nonlocal diffusions
- Pathwise stability of likelihood estimators for diffusions via rough paths
- Robust parameter estimation using the ensemble Kalman filter
- Quantitative Convergence of the Filter Solution for Multiple Timescale Nonlinear Systems with Coarse-Grain Correlated Noise
- Analysis of the feedback particle filter with diffusion map based approximation of the gain