Subsampling in ensemble Kalman inversion
arXiv:2302.11323 · doi:10.1088/1361-6420/ace64b
Abstract
We consider the Ensemble Kalman Inversion which has been recently introduced as an efficient, gradient-free optimisation method to estimate unknown parameters in an inverse setting. In the case of large data sets, the Ensemble Kalman Inversion becomes computationally infeasible as the data misfit needs to be evaluated for each particle in each iteration. Here, randomised algorithms like stochastic gradient descent have been demonstrated to successfully overcome this issue by using only a random subset of the data in each iteration, so-called subsampling techniques. Based on a recent analysis of a continuous-time representation of stochastic gradient methods, we propose, analyse, and apply subsampling-techniques within Ensemble Kalman Inversion. Indeed, we propose two different subsampling techniques: either every particle observes the same data subset (single subsampling) or every particle observes a different data subset (batch subsampling).
References in corpus (14)
- The Loss Surfaces of Multilayer Networks
- The Ensemble Kalman Filter for Inverse Problems
- Ensemble Kalman Inversion: A Derivative-Free Technique For Machine Learning Tasks
- A regularizing iterative ensemble Kalman method for PDE-constrained inverse problems
- Quantitative ergodicity for some switched dynamical systems
- Nonlinear stability and ergodicity of ensemble based Kalman filters
- Well Posedness and Convergence Analysis of the Ensemble Kalman Inversion
- A mollified Ensemble Kalman filter
- Ensemble Kalman Inversion: mean-field limit and convergence analysis
- Stochastic Modified Equations and Dynamics of Stochastic Gradient Algorithms I: Mathematical Foundations
- Analysis of Stochastic Gradient Descent in Continuous Time
- Adaptive Tikhonov strategies for stochastic ensemble Kalman inversion
- Ensemble Kalman Methods: A Mean Field Perspective
- A Continuous-time Stochastic Gradient Descent Method for Continuous Data