Optimization and Control · Cross-listed from cs.GT
Asymmetric Nash Seeking via Best Response Maps: Global Linear Convergence and Robustness to Inexact Reaction Models
Mahdis Rabbani, Navid Mojahed and Shima Nazari
Nash
equilibria provide a principled framework for modeling interactions in multi-
agent decision-making and control. However, many
equilibrium-seeking methods implicitly assume that each
agent has access to the other
agents' objectives and constraints, an assumption that is often unrealistic in practice. This letter studies a class of asymmetric-information two-player constrained games with decoupled feasible sets, in which Player 1 knows its own objective and constraints while Player 2 is available only through a best-response map. For this class of games, we propose an asymmetric projected gradient descent-best response iteration that does not require full mutual knowledge of both players' optimization problems. Under suitable regularity conditions, we establish the existence and uniqueness of the Nash
equilibrium and prove global linear convergence of the proposed iteration when the best-response map is exact. Recognizing that best-response maps are often learned or estimated, we further analyze the inexact case and show that, when the approximation error is uniformly bounded by $\varepsilon$, the iterates enter an explicit $O(\varepsilon)$ neighborhood of the true Nash
equilibrium. Numerical results on a
benchmark game corroborate the predicted convergence behavior and error scaling.
Read the full abstract ▾
Optimization and Control · Cross-listed from physics.flu-dyn
Optimization-Embedded Active Multi-Fidelity Surrogate Learning for Multi-Condition Airfoil Shape Optimization
Isaac Robledo, Alberto Vilariño, Arnau Miró et al. (+3 more)
Active multi-fidelity surrogate modeling is developed for multi-condition airfoil shape optimization to reduce high-fidelity CFD cost while retaining RANS-consistent aerodynamic metrics. The framework couples a low-fidelity-informed Gaussian process regression transfer model with uncertainty-triggered sampling and a synchronized elitism rule embedded in a hybrid genetic
algorithm. Low-fidelity XFOIL evaluations provide inexpensive features, while sparse RANS simulations are adaptively allocated when predictive uncertainty exceeds a threshold; elite candidates are mandatorily validated at high fidelity, and the population is re-evaluated to prevent evolutionary selection based on outdated fitness values produced by earlier surrogate states. The method is demonstrated for a two-point problem at $Re=6\times10^6$ with cruise at $\alpha=2^\circ$ (maximize $E=L/D$) and take-off at $\alpha=10^\circ$ (maximize $C_L$) using a 12-parameter CST representation. Independent multi-fidelity surrogates per flight condition enable decoupled refinement. The optimized design improves cruise efficiency by 41.05% and take-off lift by 20.75% relative to the best first-generation individual. Over the full campaign, RANS evaluations were required for only 14.78% and 9.5% of the condition-specific candidate evaluations at cruise and take-off, respectively. These percentages quantify the reduction in high-fidelity usage relative to the fixed automated RANS workflow adopted as the high-fidelity reference in this study.
Read the full abstract ▾
All authors Isaac Robledo, Alberto Vilariño, Arnau Miró, Oriol Lehmkuhl, Carlos Sanmiguel Vila, Rodrigo Castellanos
Optimization and Control
Stronger constraints for smooth min-max games
Valery Krivchenko, Alexander Gasnikov and Dmitry Kovalev
Saddle point problems with smooth convex-concave objective functions are often used to model min-max problems arising in machine learning. First-order methods are the standard paradigm for solving such problems. Therefore, it is important to know how those methods behave in the worst-case scenarios. In order to derive the guarantees, one would require the inequalities that appropriately constrain the iterates, gradients and function values. In this paper, we present stronger constraints for smooth convex-concave functions and show that they could allow tighter upper bounds for first-order methods.
Read the full abstract ▾
Information Theory · Cross-listed from physics.optics
Attractor-Keyed Memory
Natalia G. Berloff
Physical selectors (
lasers choosing a mode, Ising machines settling on a ground state, condensates occupying a
spin state) produce high-dimensional signatures at the moment of decision: full field amplitudes, multimode interference patterns, or scattering responses. These signatures are richer than the winner's index, yet they are routinely discarded. We show that when the signatures are repeatable across trials (stereotyped) and linearly independent across routes, a single linear decoder compiled from calibration data maps them to arbitrary payloads, merging selection and
memory access into one event and eliminating the fetch that dominates latency and energy in sparse routing architectures. The construction requires one SVD of measured device responses, which certifies capability and bounds worst-case error for any downstream payload before the task is chosen. Runtime error separates into two independently diagnosable channels, decoding fidelity (controlled by dictionary conditioning $\sigma_{\min}(\Phi)$) and routing reliability (controlled by the margin-to-noise ratio $\Delta/T_{\mathrm{eff}}$), each with a distinct physical origin and targeted remedy. We derive the full error decomposition, give Ising-machine selector constructions, and validate the predicted scalings on synthetic speckle-signature simulations across three measurement modalities. No hardware demonstration exists; we provide a falsifiable four-step experimental protocol specifying what a first experiment must measure. Whether real device signatures satisfy stereotypy is the central open question.
Read the full abstract ▾
Analysis of PDEs
Maximal Sobolev regularity of the stress tensor for the symmetric gradient p-Laplace system
Linus Behn, Andrea Cianchi, Lars Diening et al. (+1 more)
The symmetric $p$-Laplace operator enters various models in mathematical physics, such as incompressible materials with power-type hardening and non-Newtonian fluids. In this work, second-order differentiability properties of solutions to the symmetric $p$-Laplace system are established. They are formulated as maximal Sobolev regularity of the nonlinear stress tensor for locally square integrable right-hand sides.
Read the full abstract ▾
All authors Linus Behn, Andrea Cianchi, Lars Diening, Fa Peng
Optimization and Control
From Non-Identifiability to Goal-Integrated Decision-Making in Parametric Inverse Optimization
Farzin Ahmadi, Fardin Ganjkhanloo and Kimia Ghobadi
Inverse optimization seeks to recover unknown objective parameters from observed decisions, yet fundamental questions about when recovery is possible have received limited formal treatment. This paper develops a comprehensive theoretical framework for inverse optimization in parametric convex models. We first establish that non-identifiability is the generic case: even with normalization and multiple observations, the parameter set compatible with data is generically multi-dimensional, and regularization does not resolve this. We derive necessary and sufficient conditions for identifiability. Motivated by these negative results, we introduce the Inverse Learning (IL) framework, which shifts the inferential target from the unknown parameter to the latent optimal solution, achieving a complexity reduction that is independent of the number of observations. IL explicitly characterizes the full set of compatible parameters rather than returning an arbitrary element. To address the tension between observational fidelity and constraint adherence, we formalize the Observation-Constraint Tradeoff and develop Goal-Integrated Inverse Learning models that enable structured
navigation of this spectrum with guaranteed monotonicity. Numerical experiments demonstrate superior solution accuracy, higher parameter recovery rates, and significant computational speedups. We apply the framework to personalized dietary recommendations using NHANES data, proof-of-concept demonstrating improved glycemic control in a prospective feasibility study.
Read the full abstract ▾
Combinatorics
A family of Neumaier graphs containing examples with exactly five eigenvalues
Bart De Bruyn, Rhys J. Evans, Sergey Goryainov et al. (+1 more)
A Neumaier graph is an edge-regular graph with a regular clique. Such a graph is said to have parameters $(v,k,\lambda;e,s)$ if it is a $k$-regular graph on $v$ vertices having a clique of size $s$ such that every edge is contained in $\lambda$ triangles and every vertex outside $C$ is adjacent with exactly $e$ vertices inside $C$. It was an open problem whether Neumaier graphs can exist with exactly five
eigenvalues. In the present paper, we describe a family of Neumaier graphs, and show that inside this family there are 1063 nonisomorphic Neumaier graphs with parameters $(v,k,\lambda;e,s)=(48,14,2;1,4)$, among which 25 have exactly five
eigenvalues. These 1063 graphs are also the first known examples of Neumaier graphs for the mentioned parameters.
Read the full abstract ▾
All authors Bart De Bruyn, Rhys J. Evans, Sergey Goryainov, Jack Koolen
Differential Geometry
ALG gravitational instantons and Hitchin moduli spaces, I: Torelli parameters
Laura Fredrickson, Rafe Mazzeo, Jan Swoboda et al. (+1 more)
This is the first of two papers which together prove that the $12$-parameter family of
parabolic $SU(2)$-Hitchin
moduli spaces on the four-punctured sphere are all ALG gravitational instantons of type D4, and hence are asymptotic to $(\mathbb{C} \times T^2_\tau)/\mathbb{Z}_2$ at infinity. The
elliptic modulus $\tau$ is determined by the cross-ratio of the four points. In this first paper, we consider each Hitchin
moduli space corresponding to an allowable set of
parabolic data and compute its Torelli parameters. There is a $12$-parameter family of Hitchin
moduli spaces corresponding to different
parabolic data, and we show that these realize all possible allowable Torelli parameters. In the companion paper, we we will show there that all of the Hitchin
moduli spaces studied here are indeed ALG of type $D_4$, and consequently that every ALG-$D_4$ gravitational instanton can be realized as a Hitchin
moduli space. Altogether, this will give the first verification of any case of the Modularity
Conjecture: that all ALG gravitational instantons with tangent cone $\mathbb{C}/\mathbb{Z}_2$ can be realized as Hitchin
moduli spaces with their natural associated $L^2$ metrics.
Read the full abstract ▾
All authors Laura Fredrickson, Rafe Mazzeo, Jan Swoboda, Hartmut Weiss
Mathematical Physics · Cross-listed from cond-mat.stat-mech
On deforming and breaking integrability
Ysla F. Adans, Marius de Leeuw and Tristan McLoughlin
In this paper we study nearest-neighbour deformations of integrable models. After expanding in the deformation parameter, we identify four possible types of deformations. First there are deformations that simply break or preserve integrability. Then we find two different subtle cases. The first case is where the deformation is only integrable if all orders of the deformation parameter are taken into account. An example of these are the long-range deformations that appear in
holographic models. The second case is when the deformation is perturbatively integrable to some order in the deformation parameter but can not be extended to an integrable model. In this paper we work this out for the XXZ
spin chain and discuss the level statistics of each of these cases. We find numerical evidence that the onset of chaos occurs differently in each of these models. For the perturbatively integrable models, we find that the deformation strength at which chaos appears demonstrates a volume-scaling intermediate between strong and weak integrability breaking models.
Read the full abstract ▾
Numerical Analysis
A space-time dual-pairing summation-by-parts framework for forward and adjoint wave equations
Kenny Wiratama, Kenneth Duru and Yunho Kim
In this paper, we propose the first of its kind space-time dual-pairing summation by parts (DP-SBP) numerical framework for forward and adjoint wave propagation problems. This novel approach enables us to achieve spatial and temporal high order accuracy while naturally introducing dissipation in time. Within this framework, initial and boundary conditions are weakly imposed using the simultaneous approximation term (SAT) technique. Fully discrete energy estimates are derived, ensuring the stability of the resulting numerical scheme. Furthermore, the proposed space-time numerical framework allows us to construct adjoint consistent fully discrete numerical approximations, which can be applied to solve inverse wave propagation problems. We provide numerical experiments in one and two spatial dimensions to verify the theoretical analysis and demonstrate convergence of numerical errors.
Read the full abstract ▾
Geometric Topology
Exotic structures on 4-manifolds with infinite dihedral fundamental group
Simone Tagliente
The aim of this paper is to produce infinite exotic structures on smooth closed oriented $4-$
manifolds with fundamental group isomorphic to the infinite dihedral group, assuming that $b_2^+$ and $b_2^-$ are at least $12$.
Information Theory · Cross-listed from quant-ph
Tumula information and doubly minimized Petz Renyi lautum information
Lukas Schmitt, Filippo Girardi and Laura Burri
We study a doubly minimized variant of the lautum information - a reversed analogue of the mutual information - defined as the minimum relative
entropy between any product state and a fixed bipartite
quantum state; we refer to this measure as the tumula information. In addition, we introduce the corresponding Petz Renyi version, which we call the doubly minimized Petz Renyi lautum information (PRLI). We derive several general properties of these correlation measures and provide an operational interpretation in the context of hypothesis testing. Specifically, we show that the reverse direct exponent of certain binary
quantum state discrimination problems is quantified by the doubly minimized PRLI of order $\alpha\in (0,1/2)$, and that the Sanov exponent is determined by the tumula information. Furthermore, we investigate the extension of the tumula information to channels and compare its properties with previous results on the channel umlaut information [Girardi et al., arXiv:2503.21479].
Read the full abstract ▾
Optimization and Control
Constricting Tubes for Prescribed-Time Safe Control
Darshan Gadginmath, Ahmed Allibhoy and Fabio Pasqualetti
We propose a constricting Control Barrier Function (CBF) framework for prescribed-time control of control-affine systems with input constraints. Given a system starting outside a target safe set, we construct a time-varying safety tube that shrinks from a relaxed set containing the initial condition to the target set at a user-specified deadline. Any controller rendering this tube forward invariant guarantees prescribed-time recovery by construction. The constriction schedule is bounded and tunable by design, in contrast to prescribed-time methods where control effort diverges near the deadline. Feasibilityå under input constraints reduces to a single verifiable condition on the constriction rate, yielding a closed-form minimum recovery time as a function of control authority and initial violation. The framework imposes a single affine constraint per timestep regardless of state dimension, scaling to settings where grid-based reachability methods are intractable. We validate on an 18-dimensional multi-
agent system, demonstrating scalability and prescribed-time recovery with bounded control effort.
Read the full abstract ▾
Mathematical Physics · Cross-listed from gr-qc
The Structure of the Continuum Limit of Spin Foams
Matteo Bruno, Eugenia Colafranceschi, Fabio M. Mele et al. (+1 more)
The
spin foam approach to
quantum gravity aims at providing a covariant path-integral formulation of canonical loop
quantum gravity. Since
spin foam amplitudes are defined through discretizations of spacetime, understanding the continuum limit of the theory remains a central open problem. In this work, we investigate the structural aspects of this limit in a model-independent manner. We introduce an axiomatic framework for
spin foam amplitudes inspired by Atiyah's formulation of topological
quantum field theories (TQFTs). In this setting,
Hilbert spaces and amplitudes are assigned to combinatorial and topological data associated with triangulated
manifolds. By equipping the set of triangulations with suitable orders, this framework provides a precise notion of continuum limit and allows us to analyse its properties independently of any specific model. We systematically investigate how the specifics of the limit procedure allow to go beyond TQFT in the continuum. Under natural assumptions on the convergence of
spin foam amplitudes, we establish a no-go result: sufficiently strong notions of convergence necessarily lead to a topological theory. Motivated by this obstruction, we weaken the notion of convergence and consider the continuum limit of
spin foam amplitudes in a distributional sense, in the spirit of refined algebraic quantization. Under this assumption, the amplitude associated with the cylinder defines a rigging map, yielding a canonical construction of the physical
Hilbert space. The resulting continuum amplitudes act as well-defined distributions on this space of physical states, characterizing this formulation of the gravitational path integral as physical in a precise sense.
Read the full abstract ▾
All authors Matteo Bruno, Eugenia Colafranceschi, Fabio M. Mele, Carlo Rovelli
General Mathematics
On the structure of the Gram matrix for Gabor systems generated by B-splines
Martin Buck, Christina Frederick, Kasso Okoudjou et al. (+1 more)
We consider the Gabor system $\mathcal{G}(g,a\mathbb{Z}\times b\mathbb{Z})$ generated by a continuous, compactly supported function $g$ over the time-frequency
lattice generated by the parameters $a$ and $b$. We show that, under an appropriate ordering of the Gabor elements, certain submatrices of the Gram
matrix of $\mathcal{G}(g,a\mathbb{Z}\times b\mathbb{Z})$ exhibit a block-Toeplitz structure. This structural property enables us to derive spectral results for finite sub-blocks of the Gram
matrix by appealing to the spectral theory of Toeplitz
matrices. In particular, we apply our results to the Gram
matrix of Gabor systems generated by the $N$th-order B-spline.
Read the full abstract ▾
All authors Martin Buck, Christina Frederick, Kasso Okoudjou, Alexander Stangl
Dynamical Systems · Cross-listed from astro-ph.IM
Trajectory Stability and Signature Diagnostics for Comet-Based Interstellar Navigation
Bo Pieter Johannes Andrée
Interstellar objects (ISOs) motivate a coupled mission-design and inference question relevant to spacecraft dynamics and control in extreme environments: if volatile-rich, rotating comet-like bodies were used for sustained deep-space
navigation by exploiting pre-existing
hyperbolic motion and in-situ propellant, what stability requirements arise under non-gravitational forcing, and what astrometric signatures might distinguish active stabilization from uncontrolled natural dynamics? We develop a stability-theoretic framework for trajectory tracking with jet-actuated correction, and show that high-speed transit geometry – including debris-belt avoidance and encounter phasing – tightly constrains feasible trajectories, making long-horizon tracking stability mission-critical. We model tracking residuals as the balance of disturbances and corrective action, and derive stability conditions across four levels: disturbance-energy stability, outer-loop contraction, actuator-
memory stability, and rotation-mediated (Floquet) stability. The analysis implies residual diagnostics that can motivate empirical tests: under comparable forcing, effective stabilization is expected to strengthen short-horizon error correction, reduce event-conditioned persistence and variance clustering, regularize standardized innovations, and yield bounded post-shock recovery. More broadly, the framework provides a reference for deep-space guidance and control under nonlinear, multi-field disturbances and for planetary-defense concepts involving attitude shaping or impulsive kinetic impact.
Read the full abstract ▾
Optimization and Control · Cross-listed from astro-ph.IM
A Minimal Four-Thruster System for Comet-Based Interstellar Navigation
Bo Pieter Johannes Andrée
Interstellar comets arrive with key ingredients for deep-space platforms already in place: volatile inventories convertible to propellant, natural rotation providing continuous attitude variation, and
hyperbolic trajectories that carry them through the inner
Solar System and back out to interstellar space. Rather than constructing spacecraft from scratch, we ask what minimal modification is required to steer such a body along a controlled trajectory. The answer is surprisingly modest. By relaxing full six-degree-of-freedom control to forward-cone steering – sufficient for practical
navigation – we show that four thrusters suffice: one primary jet and three secondary jets at $120^\circ$ intervals. The secondary jets synthesize continuous in-plane steering, while the primary jet provides low-bandwidth attitude shaping: as the body rotates, the primary-jet torque direction sweeps predictably over a cycle, enabling out-of-plane steering via phase-scheduled firing. We formalize reachability under bounded-
curvature constraints, characterize the rotation-mediated steering envelope, discuss enabling requirements including non-
solar power at large heliocentric distances, and identify operational regimes and observable signatures implied by active trajectory control. The setting of a nutating axis is briefly considered and conjectured to preserve core results. The findings contribute to the broader effort of understanding the dynamics and control of small-body missions and offer a reference architecture relevant to long-horizon deep-space exploration and to potential planetary-defense concepts.
Read the full abstract ▾
Numerical Analysis
Interpretable AI-Assisted Early Reliability Prediction for a Two-Parameter Parallel Root-Finding Scheme
Bruno Carpentieri, Andrei Velichko, Mudassir Shams et al. (+1 more)
We propose an interpretable AI-assisted reliability diagnostic framework for parameterized root-finding schemes based on kNN-LLE proxy stability profiling and multi-horizon early prediction. The approach augments a numerical solver with a lightweight predictive layer that estimates solver reliability from short prefixes of iteration dynamics, enabling early identification of stable and unstable parameter regimes. For each configuration in the parameter space, raw and smoothed proxy profiles of a largest Lyapunov exponent (LLE)
estimator are constructed, from which contractivity-based reliability scores summarizing finite-time convergence are derived. Machine learning models predict the reliability score from early segments of the proxy profile, allowing the framework to determine when solver dynamics become diagnostically informative. Experiments on a two-parameter parallel root-finding scheme show reliable prediction after only a few iterations: the best models achieve R^2=0.48 at horizon T=1, improve to R^2=0.67 by T=3, and exceed R^2=0.89 before the characteristic minimum-location scale of the stability profile. Prediction accuracy increases to R^2=0.96 at larger horizons, with mean absolute errors around 0.03, while inference costs remain negligible (microseconds per sample). The framework provides interpretable stability indicators and supports early decisions during solver execution, such as continuing, restarting, or adjusting parameters.
Read the full abstract ▾
All authors Bruno Carpentieri, Andrei Velichko, Mudassir Shams, Paola Lecca
General Mathematics
Median-Extremes Alternation
David Carr
Given a finite linearly ordered set, we pair positions that are mirror images about its midpoint into bilateral shells, ordered by their distance from the center. Median-Extremes Alternation (MEA) begins with the central shell and repeatedly selects the unvisited shell having greatest radial contrast with the shell most recently engaged. We prove that this local rule has a strict unique maximizer at every step and forces the shell order 0, q, 1, q-1, 2, q-2, and so on, for both odd and even cardinalities. Parity affects only whether the central shell is a singleton or a pair. A separately supplied global orientation determines the order of the elements within every shell, yielding exactly two mirror traversals for n greater than or equal to 2 before orientation is fixed and exactly one afterward. Explicit formulas, a linear-time generation
algorithm, examples, and scope conditions are given. The result replaces an earlier fixed-center distance-minimization formulation, which does not generate the canonical MEA traversal.
Read the full abstract ▾
General Mathematics
An Application of Complex Fuzzy Soft Matrices in Signal Processing
Olayemi R. Oladokun and Taiwo O. Sangodapo
In this paper, we study the concept of complex fuzzy soft
matrices. The application of complex fuzzy soft
matrices in signals and systems via the cross product of complex fuzzy soft
matrices and
Fourier transform was carried out. In this application, an
algorithm for the identification of a reference signal out of large interest signals detected by a digital receiver was presented. It was recorded that, the
Fourier transform is better because it gave a higher optimal value and as a result, there was a better reference signal $R.$
Read the full abstract ▾
Combinatorics · Cross-listed from cs.DM
On the Extension Theorem for Packing Steiner Forests
Jinghan A Zeng
We consider the problem of packing edge-disjoint Steiner forests in a graph. The input consists of a multi-graph $G=(V,E)$ and a collection of $t$ vertex subsets $S = \{S_1,S_2,\ldots,S_t\}$. A Steiner forest for $S$, also called an $S$-forest, is a forest of $G$ in which each $S_i$ is connected. In the case where $t=1$, this is the Steiner Tree packing problem. Kriesell's
conjecture postulates that $2k$-edge-connectivity of $S_1$ is sufficient to find $k$ edge-disjoint $S_1$-trees. Lau showed that $24k$-edge-connectivity suffices for the Steiner Tree packing problem, which was improved to $6.5k$ by West and Wu and $5k+4$ by Devos, McDonald and Pivotto. In his thesis, Lau asserts that for the Steiner Forest problem, if each $S_i$ is $30k$-edge-connected in $G$, then there exist $k$ edge-disjoint $S$-forests. However, Lau's proof relies on an intermediate
theorem called the Extension
Theorem, which in this paper we will demonstrate has a gap by providing a counterexample to Lau's Extension
Theorem. Furthermore, we will resolve this gap by correcting Lau's proof to show that $32k$-edge-connectivity of each $S_i$ suffices to pack $k$ $S$-forests. More careful analysis yields that $31k$-edge-connectivity of each $S_i$ is sufficient when $k \geq 8$.
Read the full abstract ▾
Optimization and Control
Prescribed-Time Distributed Generalized Nash Equilibrium Seeking
Liraz Mudrik, Isaac Kaminer, Sean Kragelund et al. (+1 more)
Safety-critical multi-
agent systems, from cooperative guidance to collision avoidance, must often reach a coordinated decision by a hard deadline rather than merely converge to one eventually. This paper proposes the first fully distributed
algorithm that solves the generalized Nash
equilibrium (GNE) problem, a non-cooperative game with shared coupling constraints and general cost coupling, at a user-prescribed time $T$ independent of initial conditions. The foundation is a centralized, prescribed-time result built on the optimization Lyapunov function framework and implemented via unnormalized Hessian-gradient feedback, chosen because, unlike the Newton and normalized Hessian-gradient realizations, it naturally splits into per-
agent computations. Distributing this feedback requires each
agent to run three coupled processes simultaneously: a prescribed-time observer of the global state, a local optimization law, and a dual-consensus mechanism that enforces the shared multipliers of the variational GNE. Their simultaneous operation is the core difficulty, as the optimization continually displaces the states the observers track, while estimation errors corrupt the gradients that drive the optimization. We resolve this coupling with a multi-rate gain schedule whose observer and dual-consensus layers contract strictly faster than the optimization layer, so that every error component vanishes exactly at $T$. A Fischer-Burmeister reformulation keeps the design projection-free while enforcing the constraints at the deadline. Numerical results for a Cournot game and a time-critical sensor-coverage problem validate the approach and demonstrate its use as a solver-in-the-loop for time-critical autonomy.
Read the full abstract ▾
All authors Liraz Mudrik, Isaac Kaminer, Sean Kragelund, Abram H. Clark
Optimization and Control · Cross-listed from eess.SY
Structured Koopman Lifted Finite Memory Identification via Truncated Grunwald Letnikov Kernels
Navid Mojahed, Mahdis Rabbani and Shima Nazari
Nonlinear hereditary systems combine nonlinear state dependence with
memory. Koopman lifting provides linear predictors for nonlinear dynamics, but standard formulations are Markovian, while unrestricted lag based extensions require progressively more fitted coefficients as the retained
memory horizon increases. This work proposes a structured Koopman finite
memory model that represents history through a shared normalized Gunwald Letnikov temporal profile, while learning how the history couples the lifted coordinates directly from data. This structure retains explicit history dependence while making the number of fitted coefficients independent of the retained
memory horizon. For fixed
memory parameters, identification remains a linear regression problem with closed-form least squares and ridge solutions. We derive deterministic bounds that distinguish structured
memory approximation, omitted history, regressor conditioning, and regularization effects, and construct an exact augmented Markovian realization for recursive prediction and stability analysis. Numerical studies on a nonlinear viscoelastic robotic system demonstrate the complementary roles of nonlinear lifting and explicit
memory, while the structured representation requires substantially fewer fitted coefficients than unrestricted lag based Koopman models.
Read the full abstract ▾
Numerical Analysis
Unifying Optimization and Dynamics to Parallelize Sequential Computation: A Guide to Parallel Newton Methods for Breaking Sequential Bottlenecks
Xavier Gonzalez
Massively parallel hardware (GPUs) and long sequence data have made parallel
algorithms essential for machine learning at scale. Yet dynamical systems, like recurrent neural networks and Markov chain Monte Carlo, were thought to suffer from sequential bottlenecks. Recent work showed that dynamical systems can in fact be parallelized across the sequence length by reframing their evaluation as a system of nonlinear equations, which can be solved with Newton's method using a parallel associative scan. However, these parallel Newton methods struggled with limitations, primarily inefficiency, instability, and lack of convergence guarantees. This thesis addresses these limitations with methodological and theoretical contributions, drawing particularly from optimization. Methodologically, we develop scalable and stable parallel Newton methods, based on quasi-Newton and trust-region approaches. The quasi-Newton methods are faster and more
memory efficient, while the trust-region approaches are significantly more stable. Theoretically, we unify many fixed-point methods into our parallel Newton framework, including Picard and Jacobi iterations. We establish a linear convergence rate for these techniques that depends on the method's approximation accuracy and stability. Moreover, we give a precise condition, rooted in dynamical stability, that characterizes when parallelization provably accelerates a dynamical system and when it cannot. Specifically, the sign of the Largest Lyapunov Exponent of a dynamical system determines whether or not parallel Newton methods converge quickly. In sum, this thesis unlocks scalable and stable methods for parallelizing sequential computation, and provides a firm theoretical basis for when such techniques will and will not work. This thesis also serves as a guide to parallel Newton methods for researchers who want to write the next chapter in this ongoing story.
Read the full abstract ▾
Mathematical Physics
Solving gravitational field equations by Wiener-Hopf matrix factorisation, and beyond
M. Cristina Câmara and Gabriel Lopes Cardoso
By viewing Einstein's field equations – reduced to two dimensions – as an integrable system, one can simultaneously obtain exact solutions to both the equations themselves and their associated Lax pair via a canonical Wiener-
Hopf factorisation of a so-called monodromy
matrix. In this article, we review this remarkable interplay between gravitational field equations, integrable systems, Riemann-
Hilbert problems, and Wiener-
Hopf factorisation theory, with particular emphasis on developments from the past decade enabled by advances in Wiener-
Hopf factorisation techniques arising from the study of singular integral equations and Toeplitz operators. Through a variety of concrete examples, we illustrate how Wiener-
Hopf factorisation yields explicit, exact solutions to the field equations of gravitational theories, and how its generalisation through a so-called $\tau$-invariance property provides a new solution-generating method. Along the way, we aim to demonstrate the importance of an interdisciplinary approach – grounded in General Relativity, Complex Analysis, and Operator Theory – for the study of gravitational field equations.
Read the full abstract ▾
Probability · Cross-listed from eess.SY
Typical models of the distribution system restoration process
Arslan Ahmad and Ian Dobson
Accurate probabilistic modeling of the power system restoration process is essential for resilience planning, operational decision-making, and realistic simulation of resilience events. In this work, we develop data-driven probabilistic models of the restoration process using outage data from four distribution utilities. We decompose restoration into three components: normalized restore time progression, total restoration duration, and the time to first restore. The Beta distribution provides the best fit for restore time progression, and the Uniform distribution is a defensible, parsimonious approximation for many events. Total duration is modeled as a heteroskedastic Lognormal process that scales superlinearly with event size. The time to first restore is well described by a Gamma model for moderate and large events. Together, these models provide an end-to-end
stochastic model for Monte Carlo simulation, probabilistic duration
forecasting, and resilience planning that moves beyond summary statistics, enabling uncertainty-aware decision support grounded in utility data.
Read the full abstract ▾
Mathematical Physics
Ergodicity in discrete-time quantum walks
Kiran Kumar and Mostafa Sabri
We undertake a detailed analysis of ergodicity for homogeneous discrete-time
quantum walks on the integer
lattice. The most significant result of our paper holds in dimension one, and gives a complete equivalence between the absolutely continuous spectrum of the
unitary operator encoding the walk, and the equidistribution of its dynamics in position space, which appears for the first time in the context of large-volume
quantum ergodicity. In higher dimensions, we give a criterion for full and partial ergodicity in terms of a finer property of the spectrum which we dub “No Repeating Graphs”, and we distinguish how strongly the equidistribution is taking place (weak convergence vs total variation). Many examples are included to illustrate the criterion and to distinguish between the types of ergodicity.
Read the full abstract ▾
Combinatorics
Nearly-polynomial inverse theorem for the U^d norm in degree d+1
Tomer Milo and Guy Moshkovitz
We prove a nearly
polynomial inverse
theorem for the Gowers $U^d$ norm, over finite fields of non-small characteristic, for
polynomials of degree $d+1$. The case of degree $d$ was very recently settled by Milićević and Randelović with a fully
polynomial bound. We moreover provide a nearly
polynomial inverse
theorem for homogeneous
polynomials of any degree smaller than $2d$. Our methods may be of independent interest, and include a refined notion of
polynomial decomposition that captures correlation with
polynomials of lower degree than classical notions do, and a new correlation lemma that improves upon similar lemmas in the literature. Additionally, we illustrate the usefulness of the new correlation lemma by using it to give an alternative proof for the aforementioned result of Milićević and Randelović.
Read the full abstract ▾
Complex Variables
Bohr phenomenon for analytic and harmonic mappings on shifted disks
Vasudevarao Allu, Raju Biswas and Rajib Mandal
The primary objective of this paper is to establish several sharp results concerning the Bohr inequality, the refined Bohr inequality, and the improved Bohr inequality for the classes of analytic functions and harmonic mappings defined on the shifted disks \[ \Omega_{\gamma}=\left\{z\in\mathbb{C}:\left|z+\frac{\gamma}{1-\gamma}\right|<\frac{1}{1-\gamma}\right\}\quad\text{for}\quad\gamma\in[0,1).\]
Read the full abstract ▾
Statistics Theory · Cross-listed from stat.ME
Refined Inference for Asymptotically Linear Estimators with Non-Negligible Second-Order Remainders
Lin Li and Pengcheng Wu
First-order semiparametric inference typically uses the same influence-function linearization to justify both centering and sandwich variance estimation. We show that these are distinct questions. In a near-boundary regime, the second-order remainder can be nearly mean zero yet large enough to contribute O(n-1) sampling variation. When plug-in influence-function error and the cross term are negligible, the usual sandwich
estimator captures only the influence-function component, and Wald intervals can undercover even when the point
estimator is nearly unbiased. We derive a finite-sample variance decomposition and show that, when the cross term is asymptotically negligible, sandwich consistency is characterized by the condition nVar(Rrem) –> 0: the sandwich consistently estimates total variance in the strong-decay regime where this holds, and materially underestimates it in the near-boundary regime where nVar(Rrem) –>cR > 0. We then study two corrections. Under a leave-one-out remainder linearization condition, the jackknife consistently estimates the total variance; under a conditional Mallows-2 bootstrap condition, the pairs bootstrap does likewise. For clustered data we distinguish within-cluster dependence from first-order influence-function mismatch. Simulations for a surrogate-assisted targeted learning
estimator in stepped-wedge cluster-randomized trials illustrate variance discrepancies and the limits of jackknife correction.
Read the full abstract ▾