arXiv subjects

Sebastian Reich

Publications and source records attributed to Sebastian Reich.

At least 19 recordsLinked to original sources

arXiv2026

Robust ensemble Kalman filtering under observation noise misspecification via diffusion score matching

We address the problem of observation noise misspecification in Bayesian filtering of dynamical systems via recent advances in generalised Bayesian inference. Mis-match in tail decay between the true data generating process and an assumed observation model, often showing via frequent outliers, can strongly impact Bayesian updates and analysis in Kalman filtering. Existing approaches often employ detect-and-delete-schemes or covariance inflation to avoid assimilation of influential instances of mis-specification. In challenging settings where the analysis updates are barely sufficient to counteract the induced forecast uncertainty, these strategies may destabilize or struggle to provide reliable uncertainty quantification. We consider a novel Kalman filter adjusting information processing in the analysis step by employing diffusion score matching for inference to obtain robustness while maintaining well-quantified uncertainties. We provide theoretical properties of the diffusion score matching Kalman filter in linear Gaussian state space systems covering conjugacy and closed form parameter update in the analysis step, robustness, covariance stability, and tuning as well as high-dimensional consistency. We derive ensemble approximations via stochastic and deterministic coupling as well as implementing localization to obtain EnKF, ESRF and LETKF varieties. We evaluate the methods in appropriate simulation studies on target-tracking, the chaotic Lorenz 63 system and the Lorenz 96 system in 40 dimensions. Our insights highlight a critical trade-off between robustness and stability in Bayesian filtering. Methods employing generalized Bayesian inference can navigate this balance and improve data assimilation in challenging environments combining non-linear dynamics and potentially non-Gaussian observation noise.

math.ST↗

arXiv2026

A Continuous-Time Ensemble Kalman-Bucy Smoother for Causal Inference and Model Discovery

Data assimilation (DA) integrates observational information with model predictions to improve state estimation in complex systems. While filtering provides the basis for online forecasts by using only past and present observations, it can exhibit delays and biases when the underlying dynamics evolve rapidly or undergo regime transitions. Smoothing, which additionally incorporates future observations, provides a natural pipeline for hindcasting and reanalysis that yields an uncertainty reduction beyond the filter. This paper introduces an ensemble Kalman-Bucy smoother (EnKBS) for continuous-time DA of nonlinear dynamical systems, where the smoother's conditional distributions are reconstructed using ensemble moments. The result is a derivative-free framework that does not require explicit computation of tangent-linear or adjoint models, which converges to the exact smoother solution at the infinite-ensemble limit for a wide class of complex systems. Incorporating standard regularization techniques for high-dimensional systems, such as covariance localization and inflation, the skill of the EnKBS is demonstrated in various important scientific problems. By integrating future observations, which reveal the underlying causal mechanisms for retrospective state updates, the EnKBS is used for Bayesian-based inference of causal relationships and their temporal influence range in a dyadic trigger-feedback model and the development of a causality-driven iterative learning algorithm that identifies the structure and recovers the hidden parameters of a nonlinear reduced-order model mimicking midlatitude atmospheric circulation. Notably, both tasks remain effective with an ensemble size of $O(10)$ under partial observations, suggesting that EnKBS can support the instantaneous discovery of high-dimensional complex systems over time.

math.NA↗

arXiv2026

A McKean-Pontrygin maximum principle for entropic-regularized optimal transport

This note outlines a mean-field approach to dynamic optimal transport problems based on the recently proposed McKean-Pontryagin maximum principle. Key aspects of the proposed methodology include i) avoidance of sampling over stochastic paths, ii) a fully variational approach leading to constrained Hamiltonian equations of motion, and iii) a unified treatment of deterministic and stochastic optimal transport problems. We also discuss connections to well-known dynamic formulations in terms of forward-backward stochastic differential equations and extensions beyond classical entropic-regularized transport problems.

math.OC↗

arXiv2026

Transdimensional Data Assimilation for dynamic model selection problems

In this paper we combine the non-linear filtering capabilities of particle filters with the transdimensional inference of the reversible-jump Markov chain Monte Carlo method for a data assimilation methodology over dynamic problems with variable dimensionality. By using transdimensional MCMC steps for the rejuvenation of the particle filter, the algorithm could change the number of state space parameters on the fly and can be applied for transdimensional data assimilation purposes. Classic inversion methodologies use pre-defined models, and only changes the individual parameter values during interpretation. This is often not feasible when the optimal model parametrization is not known a priori or when the model resolution needs to change with time. The proposed transdimensional particle filter algorithm, combines the advantages of particle filters and the transdimensional MCMC methods, and provides an easily implementable data assimilation algorithm that could tackle such problems. The methodology could also improve the computational efficiency of particle filters as it could inherently optimize the model complexity in a data-driven way. We demonstrate the capabilities of the enhanced algorithm on two simple model examples.

physics.geo-ph↗

arXiv2026

Diffusion Path Samplers via Sequential Monte Carlo

We develop diffusion-based samplers for target distributions known up to a normalising constant. To this end, we rely on the well-known diffusion path that smoothly interpolates between a simple base distribution and the target, popularised by diffusion models. We tackle the score estimation problem by developing an efficient sequential Monte Carlo sampler that evolves auxiliary variables from conditional distributions along the path, providing principled score and density estimates for time-varying distributions. To control the variance of score estimates, we further propose practical control variate schedules that incur minimal overhead. We adapt this general framework to paths induced by the Ornstein-Uhlenbeck (OU) time-reversal process, stochastic interpolants, and diffusion annealed Langevin dynamics, outlining their trade-offs. Finally, we provide theoretical guarantees and empirically demonstrate the effectiveness of our method on several synthetic and real-world datasets.

stat.ML↗

arXiv2026

Gradient-free ensemble transform methods for generalized Bayesian inference in generative models

Bayesian inference in complex generative models is often obstructed by the absence of tractable likelihoods and the infeasibility of computing gradients of high-dimensional simulators. Existing likelihood-free methods for generalized Bayesian inference typically rely on gradient-based optimization or reparameterization, which can be computationally expensive and often inapplicable to black-box simulators. To overcome these limitations, we introduce a gradient-free ensemble transform Langevin dynamics method for generalized Bayesian inference using the maximum mean discrepancy. By relying on ensemble-based covariance structures rather than simulator derivatives, the proposed method enables robust posterior approximation without requiring access to gradients of the forward model, making it applicable to a broader class of likelihood-free models. The method is affine invariant, computationally efficient, and robust to model misspecification. Through numerical experiments on well-specified chaotic dynamical systems, and misspecified generative models with contaminated data, we demonstrate that the proposed method achieves comparable or improved accuracy relative to existing gradient-based methods, while substantially reducing computational cost.

stat.AP↗

arXiv2025

Affine Invariant Langevin Dynamics for rare-event sampling

We introduce an affine invariant Langevin dynamics (ALDI) framework for the efficient estimation of rare events in nonlinear dynamical systems. Rare events are formulated as Bayesian inverse problems through a nonsmooth limit-state function whose zero level set characterises the event of interest. To overcome the nondifferentiability of this function, we propose a smooth approximation that preserves the failure set and yields a posterior distribution satisfying the small-noise limit. The resulting potential is sampled by ALDI, a (derivative-free) interacting particle system whose affine invariance allows it to adapt to the local anisotropy of the posterior. We demonstrate the performance of the method across a hierarchy of benchmarks, namely two low-dimensional examples (an algebraic problem with convex geometry and a dynamical problem of saddle-type instability) and a point-vortex model for atmospheric blockings. In all cases, ALDI concentrates near the relevant near-critical sets and provides accurate proposal distributions for self-normalised importance sampling. The framework is computationally robust, potentially gradient-free, and well-suited for complex forward models with strong geometric anisotropy. These results highlight ALDI as a promising tool for rare-event estimation in unstable regimes of dynamical systems.

math.NA↗

arXiv2025

Digital Twins: McKean-Pontryagin Control for Partially Observed Physical Twins

Optimal control for fully observed diffusion processes is well established and has led to numerous numerical implementations based on, for example, Bellman's principle, model free reinforcement learning, Pontryagin's maximum principle, and model predictive control. In contrast, much fewer algorithms are available for optimal control of partially observed processes. However, this scenario is central to the digital twin paradigm, where a physical twin is partially observed and control laws are derived based on a digital twin. In this paper, we contribute to this challenge by combining data assimilation in the form of the ensemble Kalman filter with the recently proposed McKean-Pontryagin approach to stochastic optimal control. We derive forward evolving mean-field evolution equations for states and co-states which simultaneously allow for an online assimilation of data as well as an online computation of control laws. The proposed methodology is therefore perfectly suited for real time applications of digital twins. We present numerical results for controlled Lorenz-63 and Lorenz-96 systems as well as an inverted pendulum.

math.OC↗

arXiv2025

Sampling by averaging: A multiscale approach to score estimation

We introduce a novel framework for efficient sampling from complex, unnormalised target distributions by exploiting multiscale dynamics. Traditional score-based sampling methods either rely on learned approximations of the score function or involve computationally expensive nested Markov chain Monte Carlo (MCMC) loops. In contrast, the proposed approach leverages stochastic averaging within a slow-fast system of stochastic differential equations (SDEs) to estimate intermediate scores along a diffusion path without training or inner-loop MCMC. Two algorithms are developed under this framework: MultALMC, which uses multiscale annealed Langevin dynamics, and MultCDiff, based on multiscale controlled diffusions for the reverse-time Ornstein-Uhlenbeck process. Both overdamped and underdamped variants are considered, with theoretical guarantees of convergence to the desired diffusion path. The framework is extended to handle heavy-tailed target distributions using Student's t-based noise models and tailored fast-process dynamics. Empirical results across synthetic and real-world benchmarks, including multimodal and high-dimensional distributions, demonstrate that the proposed methods are competitive with existing samplers in terms of accuracy and efficiency, without the need for learned models.

stat.CO↗

arXiv2025

On a mean-field Pontryagin minimum principle for stochastic optimal control

This paper outlines a novel extension of the classical Pontryagin minimum (maximum) principle to stochastic optimal control problems. Contrary to the well-known stochastic Pontryagin minimum principle involving forward-backward stochastic differential equations, the proposed formulation is deterministic and of mean-field type. We denote it by the McKean-Pontryagin minimum principle. The Hamiltonian structure of the proposed McKean-Pontryagin minimum principle is achieved via the introduction of a pair of auxiliary functions. A gauge freedom in the choice of one of these two functions can be used to decouple the forward and reverse time equations; hence simplifying the solution of the underlying boundary value problem. We also consider infinite horizon discounted cost optimal control problems. In this case, the mean-field formulation allows one to convert the computation of the desired optimal control law into solving a pair of forward mean-field ordinary differential equations. The McKean-Pontryagin minimum principle is tested numerically for a controlled inverted pendulum, a controlled Lorenz-63 system, and a controlled Lorenz-96 system. Although the focus is on linear-quadratic control problems, the proposed methodology is extendable to more general problems including mean-field type control formulations.

math.OC↗

arXiv2025

Localized Diffusion Models

Diffusion models are state-of-the-art tools for various generative tasks. Yet training these models involves estimating high-dimensional score functions, which in principle suffers from the curse of dimensionality. It is therefore important to understand how low-dimensional structure in the target distribution can be exploited in these models. Here we consider locality structure, which describes certain sparse conditional dependencies among the target random variables. Given some locality structure, the score function is effectively low-dimensional, so that it can be estimated by a localized neural network with significantly reduced sample complexity. This observation motivates the localized diffusion model, where a localized score matching loss is used to train the score function within a localized hypothesis space. We prove that such localization enables diffusion models to circumvent the curse of dimensionality, at the price of additional localization error. Under realistic sample size scaling, we then show both theoretically and numerically that a moderate localization radius can balance the statistical and localization errors, yielding better overall performance. Localized structure also facilitates parallel training, making localized diffusion models potentially more efficient for large-scale applications.

cs.LG↗

arXiv2025

Ensemble Kalman filter for uncertainty in human language comprehension

Artificial neural networks (ANNs) are widely used in modeling sentence processing but often exhibit deterministic behavior, contrasting with human sentence comprehension, which manages uncertainty during ambiguous or unexpected inputs. This is exemplified by reversal anomalies-sentences with unexpected role reversals that challenge syntax and semantics-highlighting the limitations of traditional ANN models, such as the Sentence Gestalt (SG) Model. To address these limitations, we propose a Bayesian framework for sentence comprehension, applying an extension of the ensemble Kalman filter (EnKF) for Bayesian inference to quantify uncertainty. By framing language comprehension as a Bayesian inverse problem, this approach enhances the SG model's ability to reflect human sentence processing with respect to the representation of uncertainty. Numerical experiments and comparisons with maximum likelihood estimation (MLE) demonstrate that Bayesian methods improve uncertainty representation, enabling the model to better approximate human cognitive processing when dealing with linguistic ambiguities.

cs.CL↗

arXiv2025

Ensemble Kalman-Bucy filtering for nonlinear model predictive control

We consider the problem of optimal control for partially observed dynamical systems. Despite its prevalence in practical applications, there are still very few algorithms available, which take uncertainties in the current state estimates and future observations into account. In other words, most current approaches separate state estimation from the optimal control problem. In this paper, we extend the popular ensemble Kalman filter to receding horizon optimal control problems in the spirit of nonlinear model predictive control. We provide an interacting particle approximation to the forward-backward stochastic differential equations arising from Pontryagin's maximum principle with the forward stochastic differential equation provided by the time-continuous ensemble Kalman-Bucy filter equations. The receding horizon control laws are approximated as linear and are continuously updated as in nonlinear model predictive control. We illustrate the performance of the proposed methodology for an inverted pendulum example.

math.OC↗

arXiv2024

Localized Schr\"odinger Bridge Sampler

We consider the problem of sampling from an unknown distribution for which only a sufficiently large number of training samples are available. In this paper, we build on previous work combining Schr\"odinger bridges and plug & play Langevin samplers. A key bottleneck of these approaches is the exponential dependence of the required training samples on the dimension, $d$, of the ambient state space. We propose a localization strategy which exploits conditional independence of conditional expectation values. Localization thus replaces a single high-dimensional Schr\"odinger bridge problem by $d$ low-dimensional Schr\"odinger bridge problems over the available training samples. In this context, a connection to multi-head self attention transformer architectures is established. As for the original Schr\"odinger bridge sampling approach, the localized sampler is stable and geometric ergodic. The sampler also naturally extends to conditional sampling and to Bayesian inference. We demonstrate the performance of our proposed scheme through experiments on a high-dimensional Gaussian problem, on a temporal stochastic process, and on a stochastic subgrid-scale parametrization conditional sampling problem. We also extend the idea of localization to plug & play Langevin samplers using kernel-based denoising in combination with Tweedie's formula.

stat.ML↗

arXiv2024

Efficient, Multimodal, and Derivative-Free Bayesian Inference With Fisher-Rao Gradient Flows

In this paper, we study efficient approximate sampling for probability distributions known up to normalization constants. We specifically focus on a problem class arising in Bayesian inference for large-scale inverse problems in science and engineering applications. The computational challenges we address with the proposed methodology are: (i) the need for repeated evaluations of expensive forward models; (ii) the potential existence of multiple modes; and (iii) the fact that gradient of, or adjoint solver for, the forward model might not be feasible. While existing Bayesian inference methods meet some of these challenges individually, we propose a framework that tackles all three systematically. Our approach builds upon the Fisher-Rao gradient flow in probability space, yielding a dynamical system for probability densities that converges towards the target distribution at a uniform exponential rate. This rapid convergence is advantageous for the computational burden outlined in (i). We apply Gaussian mixture approximations with operator splitting techniques to simulate the flow numerically; the resulting approximation can capture multiple modes thus addressing (ii). Furthermore, we employ the Kalman methodology to facilitate a derivative-free update of these Gaussian components and their respective weights, addressing the issue in (iii). The proposed methodology results in an efficient derivative-free sampler flexible enough to handle multi-modal distributions: Gaussian Mixture Kalman Inversion (GMKI). The effectiveness of GMKI is demonstrated both theoretically and numerically in several experiments with multimodal target distributions, including proof-of-concept and two-dimensional examples, as well as a large-scale application: recovering the Navier-Stokes initial condition from solution data at positive times.

cs.LG↗

arXiv2024

Parameter estimation for partially observed second-order diffusion processes

Estimating parameters of a diffusion process given continuous-time observations of the process via maximum likelihood approaches or, online, via stochastic gradient descent or Kalman filter formulations constitutes a well-established research area. It has also been established previously that these techniques are, in general, not robust to perturbations in the data in the form of temporal correlations of the driving noise. While the subject is relatively well understood and appropriate modifications have been suggested in the context of multi-scale diffusion processes and their reduced model equations, we consider here an alternative but related setting where a diffusion process in positions and velocities is only observed via its positions. In this note, we propose a simple modification to standard stochastic gradient descent and Kalman filter formulations, which eliminates the arising systematic estimation biases. The modification can be extended to standard maximum likelihood approaches and avoids computation of previously proposed correction terms.

stat.ME↗

arXiv2024

Early Stopping for Ensemble Kalman-Bucy Inversion

Bayesian linear inverse problems aim to recover an unknown signal from noisy observations, incorporating prior knowledge. This paper analyses a data-dependent method to choose the scale parameter of a Gaussian prior. The method we study arises from early stopping methods, which have been successfully applied to a range of problems, such as statistical inverse problems, in the frequentist setting. These results are extended to the Bayesian setting. We study the use of a discrepancy-based stopping rule in the setting of random noise, which allows for adaptation. Our proposed stopping rule results in optimal rates for the reparameterized problem under certain conditions on the prior covariance operator. We furthermore derive for which class of signals this method is adaptive. It is also shown that the associated posterior contracts at the same rate as the MAP estimator and provides a conservative measure of uncertainty. We implement the proposed stopping rule using the continuous-time ensemble Kalman--Bucy filter (EnKBF). The fictitious time parameter replaces the scale parameter, and the ensemble size is appropriately adjusted in order not to lose the statistical optimality of the computed estimator. With this Monte Carlo algorithm, we extend our results numerically to a nonlinear problem.

math.ST↗

arXiv2024

Stable generative modeling using Schr\"odinger bridges

We consider the problem of sampling from an unknown distribution for which only a sufficiently large number of training samples are available. Such settings have recently drawn considerable interest in the context of generative modelling and Bayesian inference. In this paper, we propose a generative model combining Schr\"odinger bridges and Langevin dynamics. Schr\"odinger bridges over an appropriate reversible reference process are used to approximate the conditional transition probability from the available training samples, which is then implemented in a discrete-time reversible Langevin sampler to generate new samples. By setting the kernel bandwidth in the reference process to match the time step size used in the unadjusted Langevin algorithm, our method effectively circumvents any stability issues typically associated with the time-stepping of stiff stochastic differential equations. Moreover, we introduce a novel split-step scheme, ensuring that the generated samples remain within the convex hull of the training samples. Our framework can be naturally extended to generate conditional samples and to Bayesian inference problems. We demonstrate the performance of our proposed scheme through experiments on synthetic datasets with increasing dimensions and on a stochastic subgrid-scale parametrization conditional sampling problem as well as generating sample trajectories of a dynamical system using conditional sampling.

stat.ML↗