Corpus ID: 220496269

Imposing Sparsity Within Ensemble Kalman Inversion

@article{Schneider2020ImposingSW,
  title={Imposing Sparsity Within Ensemble Kalman Inversion},
  author={Tapio Schneider and Andrew M. Stuart and Jinlong Wu},
  journal={arXiv: Optimization and Control},
  year={2020}
}
Enforcing sparse structure within learning has led to significant advances in the field of pure data-driven discovery of dynamical systems. However, such methods require access not only to time-series of the state of the dynamical system, but also the time derivative. This poses problems when dealing with data polluted by noise, or when learning stochastic systems with non-differentiable solutions. To overcome such limitations we propose a sparse learning methodology to discover the vector… Expand
Adaptive Tikhonov strategies for stochastic ensemble Kalman inversion
TLDR
This work builds upon Tikhonov EKI (TEKI), and presents three adaptive regularization schemes, which are highlighted from both the deterministic and Bayesian approaches for inverse problems, which include bilevel optimization, the MAP formulation and covariance learning. Expand
Iterative Ensemble Kalman Methods: A Unified Perspective with Some New Variants
TLDR
This paper aims to demonstrate the efforts towards in-situ applicability of EMMARM, which aims to provide real-time information about the physical and emotional impacts of infectious disease on animals and their care and treatment. Expand
Calibration and Uncertainty Quantification of Convective Parameters in an Idealized GCM
Parameters in climate models are usually calibrated manually, exploiting only small subsets of the available data. This precludes an optimal calibration and quantification of uncertainties.Expand
Ensemble gradient for learning turbulence models from indirect observations
Training data-driven turbulence models with high fidelity Reynolds stress can be impractical and recently such models have been trained with velocity and pressure measurements. For gradient-basedExpand
lp regularization for ensemble Kalman inversion
  • Yoonsang Lee
  • Computer Science, Mathematics
  • SIAM Journal on Scientific Computing
  • 2021
TLDR
This paper proposes a strategy to implement regularization for EKI to recover sparse structures in the solution, and validate the proposed approach's effectiveness and robustness through a suite of numerical experiments, including compressive sensing and subsurface flow inverse problems. Expand

References

SHOWING 1-10 OF 18 REFERENCES
Learning Stochastic Closures Using Ensemble Kalman Inversion
TLDR
The proposed methodology for the fitting of SDE models is demonstrated, first in a simulation study with a noisy Lorenz 63 model, and then in other applications, including dimension reduction starting from various deterministic chaotic systems arising in the atmospheric sciences, large-scale pattern modeling in climate dynamics, and simplified models for key observables arising in molecular dynamics. Expand
Tikhonov Regularization within Ensemble Kalman Inversion
TLDR
This work demonstrates how further regularization can be imposed, incorporating prior information about the underlying unknown, and studies how to impose Tikhonov-like Sobolev penalties in the ensemble inversion context. Expand
On the Incorporation of Box-Constraints for Ensemble Kalman Inversion
TLDR
This work proposes a new variant of the ensemble Kalman inversion to include box constraints on the unknown parameters motivated by the theory of projected preconditioned gradient flows, and discusses a complete convergence analysis for linear forward problems. Expand
Adding Constraints to Bayesian Inverse Problems
TLDR
An approach to improve parameter estimation in inverse problems by incorporating constraints in a Bayesian inference framework and extending this framework to an approximate Bayesian inferred framework in terms of the ensemble Kalman filter method, where the constraint is imposed by re-weighing the ensemble members based on the likelihood function. Expand
Introduction to Multiple Time Series Analysis
An introduction to multiple time series analysis.
TLDR
An expository account of multiple time series analysis is presented, making it possible to ascertain dynamic leading, lagging, and feedback relationships among the series to produce more efficient forecasts and to develop control schemes. Expand
Atmospheric Modeling, Data Assimilation and Predictability
  • E. Kalnay
  • Environmental Science, Computer Science
  • 2002
TLDR
This work focuses on the post processing of numerical model output to obtain station weather forecasts and the parameterizations of subgrid-scale physical processes. Expand
Stochastic Processes and their Applications
Well, someone can decide by themselves what they want to do and need to do but sometimes, that kind of person will need some stochastic processes and their applications references. People with openExpand
Inverse Problems
The leading international journal on the theory and practice of inverse problems, inverse methods and computerized inversion of data.
Additive Manufacturing Technologies: Rapid Prototyping to Direct Digital Manufacturing
Additive Manufacturing Technologies: Rapid Prototyping to Direct Digital Manufacturing deals with various aspects of joining materials to form parts. Additive Manufacturing (AM) is an automatedExpand
...
1
2
...