Variational Dimension Lifting for Robust Tracking of Nonlinear Stochastic Dynamics
Yonatan L. Ashenafi · 2026
Read Variational Dimension Lifting for Robust Tracking of Nonlinear Stochastic Dynamics by Yonatan L. Ashenafi (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-16470.txt.
Introduction
Yonatan Ashenafi lifts state space to a high size to track wild chance paths. His 2026 preprint deals with Bayes grain tracks of chance moves. Wild curves in the move maths make tracks hard to keep true. Grains spread, clump or miss the true path in such cases. Ashenafi maps the wild model to a high size straight model.
The map can be turned back, so no facts are lost. Straight Gauss form in high size is far more easy to track. Tracks then hold true through wild bends in the path. Firm tracks serve drive, flight and fund path tools. The work gives a fresh frame for hard tracks.
Catalog record
| CREATOR | Yonatan L. Ashenafi |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 37 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2601-16470
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Variational Dimension Lifting for Robust Tracking of Nonlinear Stochastic Dynamics Yonatan L. Ashenafi Abstract Nonlinear stochastic motion presents significant challenges for Bayesian particle tracking. To address this challenge, this paper proposes a framework to construct an invertible transformation that maps the nonlinear state-space model (SSM) into a higher-dimensional linear Gaussian SSM. This approach allows the application of standard linear-Gaussian inference techniques while maintaining a connection to the dynamics of the original system. The paper derives the necessary conditions for such transformations using Ito’s lemma and variational calculus, and illustrates the method on a bistable cubic motion model, radial Brownian process model, and a logistic model with multiplicative noise. Simulations confirm that the transformed linear systems, when projected back, accurately reconstruct the nonlinear dynamics and, in distinct regimes of stiffness and singularity, yield tracking accuracy competitive with conventional filters, while avoiding their structural instabilities. 1 Introduction Dynamic state estimation in nonlinear stochastic experimental systems remains a central challenge in many areas of science and engineering. In particular, with stochastic differential equations, When nonlinear drift or diffusion terms and are present, the distribution of the solution becomes non-Gaussian, making exact tracking with a Kalman filter intractable. Standard approximate filters such as the Extended Kalman Filter (EKF) and Unscented Kalman Filter (UKF) perform approximations either by linearizing the model (EKF) or by propagating a small set of points that encode only the mean and covariance of the state distribution (UKF), and thus may perform poorly when nonlinearities are strong or the system exhibits transitions between different dynamical regimes ( 8 ; 16 ) . A different strategy is to transform the representation of the dynamics. Operator-theoretic and lifting-based methods seek coordinates in which nonlinear dynamical evolution becomes linear, as in Koopman and Carleman frameworks ( 5 ; 12 ; 10 ) . Extensions to stochastic systems include stochastic Koopman operators ( 6 ; 13 ) . These methods enable linear prediction in lifted coordinates but do not, in general, guarantee invertibility of the lifting map or consistency with Itô calculus. Furthermore, they are theoretically infinite-dimensional and practical implementations require truncation, which introduces closure errors and often necessitates high-dimensional approximations. Meanwhile, Gaussian variational surrogates approximate the time-marginal density ( 2 ) , rather than representing the dynamics themselves, which can complicate sequential filtering. This work develops a framework that constructs an Itô-consistent representation of a nonlinear stochastic system in which the dynamics become linear and Gaussian in a finite-dimensional space. The central idea is to determine both the lifting map and the associated linear dynamics so that the lifted evolution satisfies the Itô differential rules, while preserving the ability to reconstruct the original state. We formulate this as a variational problem that enforces Itô-consistency while weighting approximation error according to the stationary distribution of the original process, emphasizing fidelity where the dynamics would concentrate. This paper is organized into three main parts. First, we introduce a variational formulation to construct a finite-dimensional linear–Gaussian surrogate for nonlinear SDEs. By minimizing an objective functional weighted by the system’s stationary density, we derive a coordinate transformation that is globally consistent with Itô calculus. Second, we apply this framework to three distinct stochastic processes—cubic bistable motion, radial Bessel diffusion, and Wright-Fisher logistic evolution—and analyze how the spectral properties of the optimized linear operator impact the long-term stability of the surrogate. Third, we leverage these lifted models for Bayesian particle tracking, demonstrating how that Kalman filtering in the lifted coordinates compares with EKF, UKF, particle filtering (Sequential Monte Carlo). This approach provides a practical middle ground between Gaussian approximations, operator-theoretic infinite-dimensional embeddings, and purely approximation-based filtering, yielding interpretable linear-Gaussian surrogates of nonlinear stochastic processes with identifiable regimes of validity. 2 Background Continuous-time State space models (SSMs) are widely used to describe a given system’s dynamics. Stochastic Differential Equations (SDEs) are the favored choice for constructing these models. The SSMs are given by a motion model and an observation model. For simplicity, we will have non-dimensional equations. The motion model can be given by \displaystyle d\mathbf{x}=\mathbf{f}(\mathbf{x})dt+\mathbf{G}(\mathbf{x})\mathbf{dW}_{t} (1) where \mathbf{x}(t)\in\mathbb{R}^{n} is the latent/hidden state, \mathbf{f}:\mathbb{R}^{n}\to\mathbb{R}^{n} is the drift function describing deterministic dynamics, \mathbf{G}:\mathbb{R}^{n}\to\mathbb{R}^{n\times m} is the diffusion matrix describing how noise enters the system, and \mathbf{W}_{t}\in\mathbb{R}^{m} is an m -dimensional standard Wiener process. The observation model relates the latent states to measurable quantities: \mathbf{y}(t)=\mathbf{h}(\mathbf{x}(t))+\mathbf{v}(t), (2) where \mathbf{y}(t)\in\mathbb{R}^{p} is the observation vector, \mathbf{h}:\mathbb{R}^{n}\to\mathbb{R}^{p} is the observation function, and \mathbf{v}(t) represents observation noise, often modeled as Gaussian process. These SSM models serve as the foundation for state prediction via methods like the Kalman filter, Extended Kalman filter (EKF), Unscented Kalman filter (UKF), and particle filters (Sequential Monte Carlo). The classical Kalman filter provides exact solutions when both the motion and observation models are linear and Gaussian whereas the EKF, UKF, and other advanced methods try to give good estimates by approximation approaches. They approximate the true Bayesian update of the actual SSM model. The framework in this paper offers an alternative perspective. Instead of directly approximating the nonlinear dynamics, it transforms the original state space into a finite higher-dimensional feature space where the dynamics can be represented as linear. 3 Methods We want to transform particles’ motion and/or observation models to get a linear Gaussian SSM. We start with the general system of SDEs on \mathbb{R}^{N} for the latent states \mathbf{x} . \displaystyle d\mathbf{x}=\mathbf{f}(\mathbf{x})dt+\mathbf{G}\mathbf{dW}_{x} (3) Throughout this paper, we treat the system dynamics in dimensionless units. All state variables x and time parameters t are normalized such that the characteristic length and time scales of the motion are of order unity. Consequently, physical units are omitted. We now define the transformation T:\mathbf{x}\rightarrow\mathbf{U}(\mathbf{x})=(U_{1}(\mathbf{x}),U_{2}(\mathbf{x}),...,U_{M}(\mathbf{x})) that leads to a linear Gaussian SSM given by: \displaystyle d\mathbf{U}=\mathbf{AU}dt+\mathbf{BdW}^{e}_{t} (4) Where A and B are matrices in \mathbb{R}^{MxM} . \mathbf{dW^{e}_{t}}=[\mathbf{dW_{t}},\mathbf{0}] . We also assert that the transformation T is invertible and therefore that U_{i} is a bijective function for each i . Next we use Ito’s lemma to write down the equations that govern the transformation ( 9 ) . \displaystyle(\nabla U_{i})^{T}(\mathbf{f})+\frac{1}{2}Tr[\mathbf{G}^{T}H(U_{i})\mathbf{G}]=\sum\limits_{j=1}^{M}A_{i,j}U_{j} (5) \displaystyle(\nabla U_{i})^{T}(\mathbf{G})=\mathbf{e}_{i}^{T}\mathbf{B} (6) Where H(U_{i}) is the Hessian of U_{i} . ( 5 ) represents a system of second order linear PDEs and ( 6 ) is a system of first order PDEs that is an element of an array of such systems. We assume that G is a constant orthonormal matrix. This assumption corresponds to the regime of additive isotropic noise, which is standard in overdamped Langevin models of molecular dynamics and cellular migration where thermal fluctuations are homogeneous. The Ito condition will then be \displaystyle\mathbf{J}\mathbf{f}(\mathbf{x})+\frac{1}{2}\begin{pmatrix}\operatorname{Tr}\big[G^{T}H(U_{1})G\big]\\ \vdots\\ \operatorname{Tr}\big[G^{T}H(U_{M})G\big]\end{pmatrix}=\mathbf{AU} (7) \displaystyle\mathbf{J}\mathbf{G}\mathbf{dW_{x}}=(\mathbf{BdW^{e}_{x}})^{T} (8) Where J is the Jacobian of \mathbf{U} . Define the infinitesimal generator acting on a scalar u by \mathcal{L}u(x):=f(x)u^{\prime}(x)+\tfrac{1}{2}G(x)^{2}u^{\prime\prime}(x). Let R(x;U,A):=\mathcal{L}U(x)-A\,U(x)\in\mathbb{R}^{M},\qquad S(x;U,B):=G(x)U^{\prime}(x)-B\,U(x)\in\mathbb{R}^{M_{W}}. and define the objective functional \mathcal{J}[U,A,B]\;=\;\int_{\Omega}\Big(\|R(x;U,A)\|_{2}^{2}+\|S(x;U,B)\|_{2}^{2}\Big)\,\rho(x)\,dx. (9) Here we choose \rho to be the stationary distribution of the original SDE in ( 3 ) to keep our solution in the relevant region of state space . This weighting strategy is theoretically grounded in the ergodic property of the stochastic process ( 14 ) . Under the assumption of ergodicity, the Birkhoff Ergodic Theorem establishes the equivalence between the ensemble average over the state space and the long-term time average of a single trajectory: \lim_{T\to\infty}\frac{1}{T}\int_{0}^{T}\mathcal{E}(X_{t})\,dt=\int_{\Omega}\mathcal{E}(x)\rho(x)\,dx, where \mathcal{E}(x) represents the local approximation error (the