A systematic path to non-Markovian dynamics II: Probabilistic response of nonlinear multidimensional systems to Gaussian colored noise excitation

2025-04-10T18:21:48Z

The probabilistic characterization of non-Markovian responses to nonlinear dynamical systems under colored excitation is an important issue, arising in many applications. Extending the Fokker-Planck-Kolmogorov equation, governing the first-order response probability density function (pdf), to this case is a complicated task calling for special treatment. In this work, a new pdf-evolution equation is derived for the response of nonlinear dynamical systems under additive colored Gaussian noise. The derivation is based on the Stochastic Liouville equation (SLE), transformed, by means of an extended version of the Novikov-Furutsu theorem, to an exact yet non-closed equation, involving averages over the history of the functional derivatives of the non-Markovian response with respect to the excitation. The latter are calculated exactly by means of the state-transition matrix of variational, time-varying systems. Subsequently, an approximation scheme is implemented, relying on a decomposition of the state-transition matrix in its instantaneous mean value and its fluctuation around it. By a current-time approximation to the latter, we obtain our final equation, in which the effect of the instantaneous mean value of the response is maintained, rendering it nonlinear and non-local in time. Numerical results for the response pdf are provided for a bistable Duffing oscillator, under Gaussian excitation. The pdfs obtained from the solution of the novel equation and a simpler small correlation time (SCT) pdf-evolution equation are compared to Monte Carlo (MC) simulations. The novel equation outperforms the SCT equation as the excitation correlation time increases, keeping good agreement with the MC simulations.

Phase diagram for intermittent maps

2025-04-10T15:40:05Z

We explore the phase diagram for potentials in the space of H\"older continuous functions of a given exponent and for the dynamical system generated by a Pomeau--Manneville, or intermittent, map. There is always a phase where the unique Gibbs state exhibits intermittent behavior. It is the only phase for a specific range of values of the H\"older exponent. For the remaining values of the H\"older exponent, a second phase with stationary behavior emerges. In this case, a co-dimension 1 submanifold separates the intermittent and stationary phases. It coincides with the set of potentials at which the pressure function fails to be real-analytic. We also describe the relationship between the phase transition locus, (persistent) phase transitions in temperature, and ground states.

Equivalence of Variants of Shadowing of Free Semigroup Actions

2025-04-10T14:37:56Z

We prove that for finitely generated free semigroup actions the average shadowing property, the weak asymptotic average shadowing property, the mean ergodic shadowing property, the almost asymptotic average shadowing property, the asymptotic average shadowing property and the $M_{\alpha}$-shadowing property for every $\alpha\in (0,1)$, are equivalent. This gives an affirmative answer to an open question asked in Question 10.3 [M. Kulczycki, D. Kwietniak, P. Oprocha, On almost specification and average shadowing properties, Fundamenta Mathematicae, 224 (2014)].

Lipschitz-free spaces and dual representations of group actions

2025-04-10T14:25:18Z

We study selected topics about induced actions of topological groups $G$ on Lipschitz-free spaces $\mathcal{F}(M)$ coming from isometric actions on pointed metric spaces $M$. In particular, induced dynamical $G$-systems (under weak-star topology and the dual actions) on the dual $\mathrm{Lip_0} (M)=\mathcal{F}(M)^*$ and on the bidual $\mathcal{F}(M)^{**}$. Two such natural examples are the so-called metric compactification of isometric $G$-spaces for a pointed metric space and the Gromov $G$-compactification of a bounded metric $G$-space. One of the results asserts that for every bounded stable metric $G$-space $(M,d,\mathbf{0})$ the corresponding metric $G$-compactification $\widehat{M}$ is a weakly almost periodic $G$-flow.

Geometry and Local Recovery of Global Minima of Two-layer Neural Networks at Overparameterization

2025-04-10T13:56:57Z

Under mild assumptions, we investigate the geometry of the loss landscape for two-layer neural networks in the vicinity of global minima. Utilizing novel techniques, we demonstrate: (i) how global minima with zero generalization error become geometrically separated from other global minima as the sample size grows; and (ii) the local convergence properties and rate of gradient flow dynamics. Our results indicate that two-layer neural networks can be locally recovered in the regime of overparameterization.

Designing Universal Causal Deep Learning Models: The Case of Infinite-Dimensional Dynamical Systems from Stochastic Analysis

2025-04-10T13:41:03Z

Several non-linear operators in stochastic analysis, such as solution maps to stochastic differential equations, depend on a temporal structure which is not leveraged by contemporary neural operators designed to approximate general maps between Banach space. This paper therefore proposes an operator learning solution to this open problem by introducing a deep learning model-design framework that takes suitable infinite-dimensional linear metric spaces, e.g. Banach spaces, as inputs and returns a universal \textit{sequential} deep learning model adapted to these linear geometries specialized for the approximation of operators encoding a temporal structure. We call these models \textit{Causal Neural Operators}. Our main result states that the models produced by our framework can uniformly approximate on compact sets and across arbitrarily finite-time horizons H\"older or smooth trace class operators, which causally map sequences between given linear metric spaces. Our analysis uncovers new quantitative relationships on the latent state-space dimension of Causal Neural Operators, which even have new implications for (classical) finite-dimensional Recurrent Neural Networks. In addition, our guarantees for recurrent neural networks are tighter than the available results inherited from feedforward neural networks when approximating dynamical systems between finite-dimensional spaces.

Dimensions of orthogonal projections of typical self-affine sets and measures

2025-04-10T13:37:42Z

Let $T_1,\ldots, T_m$ be a family of $d\times d$ invertible real matrices with $\|T_i\|<1/2$ for $1\leq i\leq m$. For ${\bf a}=(a_1,\ldots, a_m)\in {\Bbb R}^{md}$, let $\pi^{\bf a}\colon \Sigma=\{1,\ldots, m\}^{\Bbb N}\to {\Bbb R}^d$ denote the coding map associated with the affine IFS $\{T_ix+a_i\}_{i=1}^m$, and let $K^{\bf a}$ denote the attractor of this IFS. Let $W$ be a linear subspace of ${\Bbb R}^d$ and $P_W$ the orthogonal projection onto $W$. We show that for $\mathcal L^{md}$-a.e.~${\bf a}\in {\Bbb R}^{md}$, the Hausdorff and box-counting dimensions of $P_W(K^{\bf a})$ coincide and are determined by the zero point of a certain pressure function associated with $T_1,\ldots, T_m$ and $W$. Moreover, for every ergodic $\sigma$-invariant measure $\mu$ on $\Sigma$ and for $\mathcal L^{md}$-a.e.~${\bf a}\in {\Bbb R}^{md}$, the local dimensions of $(P_W\pi^{\bf a})_*\mu$ exist almost everywhere, here $(P_W\pi^{\bf a})_*\mu$ stands for the push-forward of $\mu$ by $P_W\pi^{\bf a}$. However, as illustrated by examples, $(P_W\pi^{\bf a})_*\mu$ may not be exact dimensional for $\mathcal L^{md}$-a.e.~${\bf a}\in {\Bbb R}^{md}$. Nevertheless, when $\mu$ is a Bernoulli product measure, or more generally, a supermultiplicative ergodic $\sigma$-invariant measure, $(P_W\pi^{\bf a})_*\mu$ is exact dimensional for $\mathcal L^{md}$-a.e.~${\bf a}\in {\Bbb R}^{md}$.

Quasi-periodic response solutions of nonlinear plate models with nonlocal energy damping

2025-04-10T12:12:43Z

Response solutions are quasi-periodic ones with the same frequency as the forcing term. The present work is devoted to constructing response solutions for $d$-dimensional nonlinear plate models with nonlocal energy damping, which are closely related to damping phenomena in flight structures. For such models, the main characteristic is that the dissipation rate depends on the energy strength. By considering a small parameter $\epsilon$ in the domain excluding the origin and imposing a small quasi-periodic forcing with a Diophantine frequency vector, we demonstrate the persistence of the corresponding response solution. We provide an alternative approach to the contraction mapping principle (cf. [7, 33]) through a combination of reduction together with the Nash--Moser iteration technique. The reason behind this approach lies in the derivative losses caused by the nonlocal nonlinearity.

Period-Doubling Cascades Invariants: Braided Routes To Chaos

2025-04-10T09:13:46Z

By a classical result of Kathleen Alligood and James Yorke we know that as we isotopically deform a map $f:ABCD\to\mathbb{R}^2$ to a Smale horseshoe map we should often expect the dynamical complexity to increase via a period--doubling route to chaos. Inspired by this fact and by how braids force the existence of complex dynamics, in this paper we introduce three topological invariants that describe the topology of period--doubling routes to chaos. As an application, we use our methods to ascribe symbolic dynamics to perturbations of the Shilnikov homoclinic scenario and to study the dynamics of the Henon map.

On the variety of solutions of 1-dimensional nonlinear eigenvalue problems

2025-04-10T08:23:38Z

Second order nonlinear eigenvalue problems are considered for which the spectrum is an interval. The boundary conditions are of Robin and Dirichlet type. The shape and the number of solutions are discussed by means of a phase plane analysis. A new type of asymmetric solutions are discovered. Some numerical illustrations are given.

A new construction of counterexamples to the bounded orbit conjecture

2025-04-10T03:14:41Z

The bounded orbit conjecture says that every homeomorphism on the plane with each of its orbits being bounded must have a fixed point. Brouwer's translation theorem asserts that the conjecture is true for orientation preserving homeomorphisms, but Boyles' counterexample shows that it is false for the orientation reversing case. In this paper, we give a more comprehensible construction of counterexamples to the conjecture. Roughly speaking, we construct an orientation reversing homeomorphisms $f$ on the square $J^2=[-1, 1]^2$ with $\omega(x, f)=\{(-1. 1), (1, 1)\}$ and $\alpha(x, f)=\{(-1. -1), (1, -1)\}$ for each $x\in (-1, 1)^2$. Then by a semi-conjugacy defined by pushing an appropriate part of $\partial J^2$ into $(-1, 1)^2$, $f$ induces a homeomorphism on the plane, which is a counterexample.

When Koopman Meets Hamilton and Jacobi

2025-04-10T00:08:14Z

In this paper, we establish a connection between the spectral theory of the Koopman operator and the solution of the Hamilton Jacobi (HJ) equation. The HJ equation occupies a central place in systems theory, and its solution is of interest in various control problems, including optimal control, robust control, and input-output analysis. A Hamiltonian dynamical system can be associated with the HJ equation and the solution of the HJ equation can be extracted from the Hamiltonian system in the form of Lagrangian submanifold. One of the main contributions of this paper is to show that the Lagrangian submanifolds can be obtained using the spectral analysis of the Koopman operator. We present two different procedures for the approximation of the HJ solution. We utilize the spectral properties of the Koopman operator associated with the uncontrolled dynamical system and Hamiltonian systems to approximate the HJ solution. We present a convex optimization-based computational framework with convergence analysis for approximating the Koopman eigenfunctions and the Lagrangian submanifolds. Our solution approach to the HJ equation using Koopman theory provides for a natural extension of results from linear systems to nonlinear systems. We demonstrate the application of this work for solving the optimal control problem. Finally, we present simulation results to validate the paper's main findings and compare them against linear quadratic regulator and Taylor series based approximation controllers.

Monotonicity of the Liouville entropy along the Ricci flow on surfaces

2025-04-09T21:29:53Z

Using geometric and microlocal methods, we show that the Liouville entropy of the geodesic flow of a closed surface of non-constant negative curvature is strictly increasing along the normalized Ricci flow. This affirmatively answers a question of Manning from 2004. More generally, we obtain an explicit formula for the derivative of the Liouville entropy along arbitrary area-preserving conformal perturbations in this setting. In addition, we show the mean root curvature, a purely geometric quantity which is a lower bound for the Liouville entropy, is also strictly increasing along the normalized Ricci flow.

Fans, phans and pans

2025-04-09T20:45:25Z

A fan is an arcwise-connected continuum, which is hereditarily unicoherent and has exactly one ramification point. Many of the known examples of fans were constructed as 1-dimensional continua that are unions of arcs which intersect in exactly one point. Borsuk proved in 1954 that each fan is a 1-dimensional continuum which is the union of arcs intersecting in exactly one point. But it is not yet known if this property is equivalent to being a fan. In this paper, we show that under two additional assumptions, every such union of arcs is a fan.

Uniform property $Γ$ and the small boundary property

2025-04-09T20:00:46Z

We prove that, for a free action $\alpha \colon G \curvearrowright X$ of a countably infinite discrete amenable group on a compact metric space, the small boundary property is implied by uniform property $\Gamma$ of the Cartan subalgebra $(C(X) \subseteq C(X) \rtimes_\alpha G)$. The reverse implication has been demonstrated by Kerr and Szab\'o for free actions, from which we obtain that these two conditions are equivalent. We moreover show that, if $\alpha$ is also minimal, then almost finiteness of $\alpha$ is implied by tracial $\mathcal{Z}$-stability of the subalgebra $(C(X) \subseteq C(X) \rtimes_\alpha G)$. The reverse implication is due to Kerr, resulting in the equivalence of these two properties as well. As an application, we prove that if $\alpha \colon G \curvearrowright X$ and $\beta \colon H \curvearrowright Y$ are free actions and $\alpha$ has the small boundary property, then $\alpha \times \beta \colon G \times H \curvearrowright X \times Y$ has the small boundary property. An analogous permanence property is obtained for almost finiteness in case $\alpha$ and $\beta$ are free minimal actions.