# Unbiased Estimation of the Vanilla and Deterministic Ensemble Kalman-Bucy Filters

@article{Alvarez2022UnbiasedEO,
title={Unbiased Estimation of the Vanilla and Deterministic Ensemble Kalman-Bucy Filters},
author={Miguel Alvarez and Neil K. Chada and Ajay Jasra},
journal={ArXiv},
year={2022},
volume={abs/2208.03947}
}
• Published 8 August 2022
• Mathematics
• ArXiv
In this article we consider the development of an unbiased estimator for the ensemble Kalman– Bucy ﬁlter (EnKBF). The EnKBF is a continuous-time ﬁltering methodology which can be viewed as a continuous-time analogue of the famous discrete-time ensemble Kalman ﬁlter. Our unbiased estimators will be motivated from recent work [Rhee & Glynn 2010, [31]] which introduces randomization as a means to produce unbiased and ﬁnite variance estimators. The randomization enters through both the level of…

## References

SHOWING 1-10 OF 36 REFERENCES

### Multilevel Estimation of Normalization Constants Using the Ensemble Kalman-Bucy Filter

• Computer Science
Stat. Comput.
• 2022
This article provides an analysis based on deriving L q -bounds for the normalizing constants using both the single-level, and the multilevel algorithms, which is largely based on previous work deriving the MLEnKBF.

### On the Mathematical Theory of Ensemble (Linear-Gaussian) Kalman-Bucy Filtering

• Computer Science
• 2020
An overview of the theory of ensemble Kalman-Bucy filtering for linear-Gaussian signal models and some non-asymptotic time-uniform fluctuation, stability, and contraction results on the sample covariance and sample mean in continuous-time ensemble filtering.

### On the Stability of Kalman-Bucy Diffusion Processes

• Mathematics
SIAM J. Control. Optim.
• 2017
The purpose of this work is to review, re-prove and refine existing results concerning the dynamical properties of the Kalman-Bucy filter so far as they pertain to filter stability and convergence, with a number of bounds, convergence, and eigenvalue inequalities rigorously proven.

### Ensemble Kalman methods for inverse problems

• Mathematics
• 2012
It is demonstrated that the ensemble Kalman method for inverse problems provides a derivative-free optimization method with comparable accuracy to that achieved by traditional least-squares approaches, and that the accuracy is of the same order of magnitude as that achieve by the best approximation.

### Multilevel Ensemble Kalman-Bucy Filters

• Mathematics, Computer Science
SIAM/ASA J. Uncertain. Quantification
• 2022
It is proved that the multilevel EnKBF can achieve a mean square error (MSE) of $\mathcal{O}(\epsilon^2), \epsil on>0$ with a cost of order with a reduction in cost compared to the (single level) En KBF, and a Monte Carlo convergence and approximation bounds associated to time-discretized EnKBFs.

### On Unbiased Score Estimation for Partially Observed Diffusions

• Mathematics, Computer Science
• 2021
This work constructs unbiased estimators of the score function, i.e. the gradient of the log-likelihood function with respect to parameters, with no time-discretization bias, and proves under assumptions that these estimators are unbiased and have finite variance.

### On Unbiased Estimation for Discretized Models

• Mathematics, Computer Science
ArXiv
• 2021
This article considers the construction of unbiased Monte Carlo estimators of expectations w.r.t. probability measures which are subject to discretization error, and shows how to obtain such estimators using a novel adaptation of randomization schemes and Markov simulation methods.

### On the stability and the uniform propagation of chaos properties of Ensemble Kalman-Bucy filters

• Mathematics
• 2016
The Ensemble Kalman filter is a sophisticated and powerful data assimilation method for filtering high dimensional problems arising in fluid mechanics and geophysical sciences. This Monte Carlo

### Unbiased Inference for Discretely Observed Hidden Markov Model Diffusions

• Mathematics, Computer Science
SIAM/ASA J. Uncertain. Quantification
• 2021
A Bayesian inference method for diffusions observed discretely and with noise, which is free of discretisation bias is developed, based on particle marginal Metropolis--Hastings, a particle filter, randomised multilevel Monte Carlo, and importance sampling type correction of approximate Markov chain Monte Carlo.

• Mathematics