Peer Revenue Play credits. Real papers.

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

Mathematics

math Watch Mathematics (sign in to use)

Friday 24 April

Algebraic Geometry

Notes on the deformed Hermitian-Yang-Mills equations and the large scaling limits of stability conditions

In this short note, we show that, assuming a conjecture of Arcara and Miles, a line bundle on a smooth complex projective surface admits a deformed Hermitian-Yang-Mills metric if and only if it is stable in the “large scaling limit" with respect to a generic Kähler form. The same statement for toric surfaces was recently proved by Stoppa. The purpose of this note is to remark that this equivalence holds for arbitrary smooth projective surfaces.
Read the full abstract

Geometric Topology

Finiteness of integral representations on 2-perfect truncation polytopes

Let $P$ be a compact hyperbolic Coxeter truncation polytope of dimension $d\ge 3$, and let $\Gamma$ be the orbifold fundamental group of the associated Coxeter orbifold $\mathcal{O}_P$. Let $\mathscr{G}(\Gamma,G)$ be the geometric component containing the holonomy representation in $\operatorname{Hom}(\Gamma,G)/G$. $\mathscr{G}(\Gamma,G)$ is identified with the deformation space of properly convex real projective structures on the Coxeter orbifold $\mathcal{O}_P$. We prove that $\mathscr{G}(\Gamma,G)$ contains only finitely many integral representations. The same conclusion holds more generally for irreducible, large, $2$-perfect truncation polytopes.
Read the full abstract

Optimization and Control

Using a generative model for out-of-sample testing of two-stage stochastic programs

Stochastic programming models for decision-making under uncertainty often suffer from scenario scarcity, where obtaining representative samples of uncertain parameters requires expensive simulations or measurements. This work presents a framework that leverages the Normal-to-Anything (NORTA) generative model to enhance the reliability of two-stage stochastic programming solutions through comprehensive out-of-sample testing when scenario data is limited. The NORTA model efficiently generates synthetic scenarios that preserve both marginal distributions and correlation structures from limited available data, offering a computationally tractable alternative to expensive physics-based simulations. We demonstrate the approach through a case study on power grid resilience planning against flood events in Texas, where we use 16 high-fidelity flood scenarios to generate 800 additional synthetic scenarios for validation. The results show that NORTA-generated scenarios accurately capture essential statistical properties, with the out-of-sample performance of first-stage decisions closely matching expectations from the original stochastic programming model. This framework enables decision-makers to assess the robustness of their solutions when obtaining additional real-world data is prohibitively expensive. The approach bridges machine learning and operations research by providing a practical solution to scenario generation challenges in stochastic programming.
Read the full abstract

Combinatorics

Jaeger-type orientations of random regular graphs

We consider $p$-orientations, which are defined to be orientations of $d$-regular graphs such that every vertex either has in-degree $p$ or out-degree $p$. These generalise the orientations considered in Jaeger's conjecture, where $d=4p+1$. Working with random $d$-regular graphs using the small subgraph conditioning method, we prove that a $d$-regular graph has a $p$-orientation with high probability for several values of $(d,p)$, including the $p=3,4$ cases of Jaeger's conjecture (known to be deterministically false). Some negative results are obtained by exploiting a connection with maximum bisection size.
Read the full abstract

Numerical Analysis

A data-driven model reduction approach for backward fractional diffusion-wave equations

In this work, we propose an observation system based on the available data which solution is one-be-one mapping to the forward problem(with the unknown initial function) solution. It implies their solutions share the same linear structure in the finite dimensional space. Theoretical results show model reduction approaches constructed for the observation system also work well for the forward problem, which significantly improve the efficiency of solving the inverse problem. Several numerical examples are presented to support our finding.
Read the full abstract

Numerical Analysis

Finite Expression Method with TranNet-based Function Learning for High-Dimensional Partial Differential Equations

In this paper, we study a machine-learning-based solver for high-dimensional partial differential equations (PDEs). Computing accurate solutions efficiently for such problems remains challenging because of the curse of dimensionality, which severely limits the scalability of classical numerical methods. Our approach builds on the recently developed finite expression method (FEX), which approximates PDE solutions in a function space generated by finitely many analytic expressions. This framework has been shown to achieve high, and in some cases machine-level, accuracy with polynomial memory complexity and favorable computational cost. We propose an extension of FEX in which the functional pool is generated by shallow neural network operators whose parameters are initialized using the transferable neural network method TransNet. Numerical experiments suggest that the proposed extension is an effective alternative for solving several high-dimensional PDEs.
Read the full abstract

All authors Phuoc-Toan Huynh, Feng Bao, Haizhao Yang, Ahmed Zytoon

Analysis of PDEs

Gradient regularity for nonlocal double phase equations

This paper is devoted to investigating the interior $C^{1, \alpha}$ regularity of viscosity solutions to the nonlocal double phase equations $$ \int_{\mathbb{R}^d} \left(\frac{|u(x)-u(y)|^{p-2}(u(x)-u(y))}{|x-y|^{d+sp}}+a(x,y)\frac{|u(x)-u(y)|^{q-2}(u(x)-u(y))}{|x-y|^{d+tq}}\right)dy=0, $$ where $2\le p\le q$, $0<s\le t<1$, and $a(x, y)\ge0$. By assuming the Lipschitz continuity of $a(\cdot)$, we show that the gradient of solution is Hölder continuous, provided the distance of $tq$ and $sp$ is suitably small. As a key ingredient to this conclusion, the Lipschitz property of solutions is also established under weaker assumptions on the modulating coefficient $a(\cdot)$, which is of independent interest. Our results develop a nonlocal counterpart of the gradient regularity theory for classical double phase problems due to Colombo & Mingione [Arch. Ration. Mech. Anal., 2015] and solve the higher regularity issue raised by De Filippis & Palatucci [J. Differential Equations, 2019]. The core challenges consist in precisely characterizing the subtle interaction among the pointwise behaviour of the coefficient $a(\cdot)$, the growth exponents and the differentiability orders.
Read the full abstract

Analysis of PDEs

Trace estimates and improved pointwise bounds for joint eigenfunctions

For $L^2$-normalized joint eigenfunctions in a quantum integrable system, [GT20] gave polynomial improvements over the standard Hömander bounds for typical points. In this paper, we improve their result by establishing a sharp bound of $h^{\frac{-n+k+1}2}$ for the points satisfying a rank $k$ non-degeneracy condition.
Read the full abstract

Category Theory

Homotopic morphisms and diagram theorems in extriangulated categories

Homotopic morphisms of $\mathbb E$-triangles in extriangulated categories are introduced. Any morphism of $\mathbb E$-triangles is a composition of homotopic morphisms. Any morphism $(\alpha_1, \alpha_2, \alpha_3)$ of $\mathbb E$-triangles can be modified to be homotopic, by changing one of $\alpha_i$; moreover, all the 15 cases where $\alpha_i$ is an $\mathbb E$-inflation ($\mathbb E$-deflation) are analyzed. Some diagram theorems, especially $4\times 4$ Lemma and its $14$ variants, including $3\times 3$ diagram and Horseshoe Lemma, are investigated. A relation between homotopic morphisms and (middling) good morphisms in triangulated categories are given. Weakly idempotent complete extriangulated categories are characterized.
Read the full abstract

Mathematical Physics · Cross-listed from quant-ph

Signature of paraparticles: a minimal Gedankenexperiment

Paraparticles beyond bosons and fermions can be exchanged via either the braid group (anyons, existing up to $D=2$ space dimensions) or the permutation group; in the latter case the space dimensions are not limited. Besides being predicted, anyons have been experimentally detected. The situation differs for paraparticles exchanged via the permutation group ("permutation-group parastatistics").The first test to detect their theoretical signature was published in 2021 (for $Z_2\times Z_2$-graded parafermions; it was soon followed by a second paper proving the detectability of $Z_2\times Z_2$-graded parabosons). Later on, two further papers proved theoretical signatures of permutation-group parastatistics. These works demonstrate that, in certain situations, a long-held belief on the "conventionality of parastatistics" argument can be evaded: some measurements of permutation-group paraparticles cannot be recovered from ordinary bosons/fermions. The main question now is how to experimentally detect or engineer in the laboratory such paraparticles. For this aim a minimal setup for the theoretical test is here provided: a Gedankenexperiment (a simplified version of the two tests published in 2021) which, essentially, is a flow chart of logical operations. The key point is to present, to experimentalists, the necessary steps to be simulated/realized in the laboratory (possibly, by manipulating qudits). In this minimal setup, the detection/engineering of paraparticles is mapped into a chirality test. The mathematical setting is based on $Z_2\times Z_2$-graded color Lie (super)algebras and derived mathematical structures.
Read the full abstract

Algebraic Geometry

Algebraic methods in periodic singular Liouville equations

We explain how algebraic geometry comes into play in the study of non-linear mean field (singular Liouville) equations $$ \triangle u + e^u = 4\pi \sum_{i = 1}^N \ell_i \delta_{p_i} $$ on a flat torus $E = \Bbb C/\Lambda$, where $N, \ell_1, \ldots, \ell_N \in \Bbb N$, $p_i \in E$ are distinct points, and $\delta_{p_i}$ is the Dirac measure at $p_i$. The case with one singular source ($N = 1$) had been studied extensively in recent years. We start with a survey of this case with emphasizes on the constructions of Lamé curves $\overline X_n$ and pre-modular forms $Z_n(\sigma, \tau)$ which encodes the structure of solutions of the PDE. We then discuss extensions to the case of general $N$. The basic tool is the monodromy theory for generalized Lamé equations. Two aspects are discussed: (1) For $\ell := \sum_{i = 1}^N \ell_i$ being odd, an exact counting formula of algebraic degree is proved. (2) For $\ell$ being even, the existence of generalized Lamé curves parametrizing logarithmic-free solutions is proposed.
Read the full abstract

Dynamical Systems

No infinite spin for total collisions in the spatial N-body problem

In the $n$-body problem, when bodies tend to a total collision, then its normalized shape curve converges to the set of normalized central configurations, which has $SO(3)$ symmetry in the planar case. This leaves a possibility that the normalized shape curve tends to the set obtained by rotations of some central configuration instead of a particular point on it. This is the infinite spin problem which concerns the rotational behavior of total collision orbits in the $n$-body problem. We show that the infinite spin is not possible if the limiting shape is isolated from other connected components of the set of normalized central configurations. Our approach extends the method from recent work for total collision for the planar case by Moeckel and Montgomery. The main tool is a full reduction $\rm SO(3)$–symmetry in a context of vanishing angular momentum.
Read the full abstract

Functional Analysis

Representer Theorem in Complex Reproducing Kernel Hilbert Spaces with Applications to Fock and Hardy Spaces and Superoscillations

We introduce a complex-valued counterpart of the representer theorem in machine learning. We study several learning and minimization problems in reproducing kernel Hilbert spaces (RKHSs), with the aim of identifying appropriate input-output data sets that allow specific functions to appear as solutions of regression-type minimization problems. In particular, we recover superoscillations in the Fock space, the Gaussian radial basis function (RBF) kernel in the corresponding RKHS, and finite Blaschke products in the Hardy space setting. We then extend the notion of superoscillations through suitable generalizations of the Fock space and investigate the associated learning problems. This is a seminal work relating superoscillations and machine learning kernel methods via the representer theorem.
Read the full abstract

Probability

Adapted Optimal Transport between Filtered Gaussian Processes

We continue the study of adapted optimal transport in the discrete-time Gaussian setting. To this end, we introduce a space of filtered Gaussian processes where both the randomness and the flow of information are driven by a Gaussian white noise. On this space, the adapted $2$-Wasserstein distance (${AW}_2$) admits a variational representation as a constrained orthogonal Procrustes problem between Cholesky factors. Furthermore, the resulting quotient space is the ${AW}_2$-completion of the space of Gaussian distributions on the path space. We also characterize explicitly the ${AW}_2$-projections onto the subspaces of Gaussian martingales. Next, we analyze the adapted Brenier coupling – a multivariate generalization of the Knothe–Rosenblatt coupling that serves as a myopic solution to the adapted transport problem, and compute its transport cost. Utilizing a Gaussian random matrix framework, we investigate the asymptotic behavior of transport costs as the time horizon grows; notably, we establish that the transport costs of all Gaussian bicausal couplings are asymptotically equivalent, whereas the classical Bures–Wasserstein distance is strictly smaller. Finally, we demonstrate that the adapted analogue of Gelbrich's lower bound fails in general, and we identify a sufficient martingale difference condition under which the bound is recovered.
Read the full abstract

Optimization and Control

Rate-Optimal Regret for the Safe Learning-based Control of the Constrained Linear Quadratic Regulator

We study the problem of adaptive control of the stochastic linear quadratic regulator (LQR) with constraints that must be satisfied at every time step. Prior work on the multidimensional problem has shown $\tilde{O}(T^{2/3})$ regret and satisfaction of robust constraints, leaving open the question of whether $\tilde{O}(\sqrt{T})$ regret can be attained in the constrained LQR setting. We contribute to this problem by showing $\tilde{O}(\sqrt{T})$ regret and satisfaction of chance constraints. This type of constraints allow us to handle unbounded noise and also enable analytical techniques not directly applicable to robust constraints. Our proposed algorithm for this problem uses an SDP to select an optimistic policy, and then "scales back" this policy until it is verifiably-safe. Our theoretical analysis establishes regret and constraint guarantees via a key lemma that bounds the system covariance in terms of the chosen policy. This covariance-based analysis is in contrast with the cost-to-go based analysis that is typically used in adaptive LQR.
Read the full abstract

Number Theory

The exceptional set for Diophantine inequality with mixed powers of primes

Assume that $\lambda_1, \lambda_2, \lambda_3,\lambda_4,\lambda_5,\lambda_6,\lambda_7$ are non-zero real numbers , $\lambda_1/\lambda_2$ is an irrational number. Let $\mathcal{V} $ be a well-spaced sequence, and $\delta >0$. For any given positive integer $k\geq 5$ and any $\varepsilon >0$, we give the upper bound of the number of $\upsilon \in \mathcal{V} $ with $\upsilon \leq X$ for which the inequality $$ \left | \lambda_1p_1^2 + \lambda_2p_2^3 + \lambda_3p_3^3 + \lambda_4p_4^3 + \lambda_5p_5^3 + \lambda_6p_6^4 + \lambda_7p_7^k - \upsilon \right | <{\upsilon}^{-\delta} $$ has no solution in primes $p_1, p_2, p_3, p_4, p_5, p_6, p_7$.
Read the full abstract

Mathematical Physics

Tetrahedral $L$-operators, tensor Schur polynomials and $q$-deformed loop elementary symmetric functions

We study three-dimensional partition functions constructed from the tetrahedral $L$-operator introduced and studied by Bazhanov-Sergeev and Kuniba-Maruyama-Okado. First, we explore the $q=0$ case, extending the authors' previous results and giving applications by a further analysis on the Zamolodchikov-Faddeev algebra. We introduce a class of partition functions which can be expressed as the tensor Schur polynomials, a class of products of Schur polynomials. As an application, we derive the shuffle formula for the Schur polynomials which is geometrically the pushforward formula by Joźefiak-Pragacz-Lascoux. We also give a derivation and a unification of the Gustafson-Milne and Fehér–Némethi–Rimányi identities, and introduce a family of Laurent polynomials using divided difference operators which imitates the Schubert polynomials from the perspective of our study. We also present an application to the steady state of the multispecies totally asymmetric simple exclusion process. Second, we investigate several classes of partition functions for the generic $q$ case, and determine the explicit forms as deformations of the elementary symmetric functions. One of them can be regarded as (an extension of) a $q$-deformed loop elementary symmetric functions.
Read the full abstract

Statistics Theory · Cross-listed from stat.ML

Concave Statistical Utility Maximization Bandits via Influence-Function Gradients

We study stochastic multi-armed bandits in which the objective is a statistical functional of the long-run reward distribution, rather than expected reward alone. Under mild continuity assumptions, we show that the infinite-horizon problem reduces to optimizing over stationary mixed policies: each weight vector \(w\) on the simplex induces a mixture law \(P^w\), and performance is measured by the concave utility \(U(w)=\mathfrak U(P^w)\). For differentiable statistical utilities, we use influence-function calculus to derive stochastic gradient estimators from bandit feedback. This leads to an entropic mirror-ascent algorithm on a truncated simplex, implemented through multiplicative-weights updates and plug-in estimates of the influence function. We establish regret bounds that separate the mirror-ascent optimization error from the bias caused by estimating the influence function. The framework is developed for general concave distributional utilities and illustrated through variance and Wasserstein objectives, with numerical experiments comparing exact and plug-in influence-function implementations.
Read the full abstract

Optimization and Control

Training Dynamics and Induced Preconditioning of ReLU Policy Gradient for Scalar LQR

Policy gradient is not invariant to reparameterization; the training dynamics of the weights of a neural controller differ from those of the induced gains. We make that difference exact for deterministic scalar discounted linear-quadratic regulation with bias-free one-hidden-layer ReLU policies. We show that policy gradient on the weights induces preconditioning on the gains' dynamics, alongside an explicit quadratic finite-step correction and an activation-group error. Moreover, we identify explicit conditions on the step sizes and initialization parameters that guarantee that the sequence of generated controllers remains stabilizing throughout training and converges to the optimal gains with high probability. For a given system and cost parameters, the sufficient step-size conditions imply contraction bounds independent of the network width. Numerical analysis illustrates how the neural realization changes the induced gain dynamics while maintaining stability across training.
Read the full abstract

Combinatorics

On 132-Avoiding Permutations with an Adjacency Constraint

We study permutations in $S_n$ that simultaneously avoid the pattern $132$ and satisfy the adjacency bound $|\pi_{i+1} - \pi_i| \leq m$ for all $i$, denoting their number by $A_n^{(m)}$. This combination of a global pattern restriction and a local bounded-difference condition produces a strong structural collapse: whereas unrestricted $132$-avoiding permutations are counted by the Catalan numbers with exponential growth rate $4$, the adjacency constraint forces the maximum element $n$ to occupy only positions in $\{1, 2, \ldots, m\} \cup \{n\}$. We give a complete solution for $m = 2$ by partitioning the class according to the position of the maximum element. This yields explicit recurrences and a rational generating function, from which we derive asymptotic growth of the form $A_n^{(2)} \sim C \alpha^n$ with $\alpha \approx 1.4656$. We conjecture that for each fixed $m$, the class admits a finite-state structural decomposition leading to linear recurrences with constant coefficients and rational generating functions, with growth constants increasing to $4$.
Read the full abstract

Algebraic Geometry

Integral Perverse Obstructions for Normal Surface Singularities: Resolution Determinants and Monodromy

For a germ $(X,0)$ of a normal complex analytic surface, let $E:=H^0({}^p_+IC_X\mathbb Z)_0$, where ${}^pIC_X\mathbb Z$ and ${}^p_+IC_X\mathbb Z$ denote the ordinary and dual middle-perversity intersection complexes with integral coefficients. This finite abelian group measures the integral discrepancy between the two middle extensions. Motivated by work of Jung–Saito, we study $E$ as a local invariant of the singularity. We prove that $E$ admits a topological realization as $H^2(L,\mathbb Z)_{\tors}$, where $L$ is the link of the singularity, and a geometric realization as the discriminant group of the exceptional lattice of the minimal resolution. In particular, if $M$ is the intersection matrix of the irreducible exceptional curves, then $|E|=|\det(M)|$. If $(X,0)$ is an isolated hypersurface surface singularity, we further prove that $E\cong \coker(T-\id)_{\tors}$, where $T$ is the Milnor monodromy on integral vanishing cohomology. Under the additional hypothesis that $(T-\id)\otimes_{\mathbb Z}\mathbb Q$ is an isomorphism, this yields $|E|=|\det(T-\id)|$. Thus the same local integral obstruction admits compatible perverse, topological, resolution-theoretic, and monodromy-theoretic realizations.
Read the full abstract

Probability

Reflected Stochastic Differential Equations Driven by G-Brownian Motion with Nonlinear Constraints

In this paper, we study the reflected stochastic differential equations driven by G-Brownian motion (reflected G-SDEs) with two nonlinear constraints. With the help of the Skorokhod problem with nonlinear constraints, we first study the doubly reflected G-Brownian motion, which is constructed pathwise and lies in the same G-expectation space as the G-Brownian motion. For the reflected G-SDE, the uniqueness is derived from some a priori estimate and the existence is obtained by a Picard iteration method. The comparison theorem of the solution and the individual constraining processes are provided.
Read the full abstract

Thursday 23 April

Optimization and Control

Comparing Optimization Models for Radiotherapy Scheduling

The Radiotherapy Scheduling Problem (RTSP) involves determining an optimal schedule for patients undergoing radiation treatments, a task that has a massive impact on clinical outcomes given the central role of radiotherapy in cancer care. The daily batch approach–which consists of scheduling all the newly arrived patients together at the end of each day–modelled with Integer Linear Programming, is currently one of the most effective methods for the RTSP. However, this kind of formulation requires substantial computational resources in terms of time and memory. Here, we address these limitations by developing two novel greedy heuristics (named RTSP First Fit and RTSP Best Fit) and use them as constructive heuristics for a Simulated Annealing (SA) approach to optimize the scheduling. The proposed methods–the heuristics alone and their combination with SA–are evaluated on a publicly available dataset against an integer linear program formulation solved with two different state-of-the-art exact solvers. Evaluation metrics include six scheduling objectives capturing patient waiting times, preference satisfaction, and changes in linear accelerator assignment (aggregated in four different weight configurations), solving time, and memory consumption. The results show that the novel heuristics achieve solutions close to those of exact methods, while dramatically reducing runtime and memory usage; furthermore, when combined with SA, they further improve the solution quality while maintaining low runtime and memory usage.
Read the full abstract

All authors C.C. Rambaldi Migliore, D. Stanicel, N. Musliu, G. Iacca, M. Roveri

General Mathematics

Robust Chance Constrained Complex Zero-Sum Games

This paper develops a unified framework for zero-sum games in which both the pure strategies and the payoff matrices contain complex-valued entries. By leveraging a linear isomorphism between complex and real vector spaces, we extend key results from real-valued convex analysis to the complex domain, establishing the validity of the minimax theorem and the preservation of saddle-point structure. Building on this foundation, we formulate a complex zero-sum game model that enables mixed strategies to interact with the real and imaginary components of the payoff matrix, and we characterize its saddle-point equilibrium through associated primal and dual problems. To incorporate uncertainty, we introduce a complex chance-constrained zero-sum game model (3CP) that handles individual probabilistic constraints defined by complex linear functionals. We first study the 3CP formulation under known exact distributions, focusing on Complex Elliptically Symmetric random variables, which generalize the complex Gaussian family. The framework is then extended to moments-based ambiguity sets, including: (i) distributions with known first two moments, (ii) distributions with unknown second-order moments, and (iii) fully distributed with unknown moments. In all cases, the probabilistic constraints admit deterministic second-order cone representations, ensuring convex feasible strategy sets and enabling explicit characterization of the complex game value. Numerical experiments, including a transmitter–jammer waveform interaction model, show how the proposed framework captures the behavior of complex mixed strategies. Additionally, we evaluate out-of-sample rates and confirm that practical behavior closely aligns with the theoretical guarantees.
Read the full abstract

Optimization and Control

Determinant Bounds for $(n-1)$-Locally Positive Semidefinite Matrices

In this framework, the extremal case corresponds to the tightest nontrivial relaxation in this hierarchy, in which every proper principal submatrix is constrained to be positive semidefinite, while the global positive semidefiniteness condition is governed by the determinant. In this paper, we study the determinants of locally positive semidefinite matrices and derive sharp lower bounds on their determinants that quantify the gap between local and global positive semidefiniteness. We further obtain analogous extensions of classical determinant inequalities, including Fisher and Koteljanskii inequalities, providing tight lower bounds in each case. In a sense, these results quantify, via determinant bounds, how far the class of locally positive semidefinite matrices can be from being positive semidefinite.
Read the full abstract

Probability · Cross-listed from q-bio.NC

Noise-Induced Transitions and Coherence Resonance in a 5D Conductance-Based Neuronal Model

Intrinsic channel noise is an important source of variability in neuronal dynamics, but its state-dependent and boundary-constrained nature can be difficult to represent numerically. Here, we investigate the effects of bounded multiplicative noise applied to the slow M-current gating variable in a five-dimensional conductance-based model of a CA1 pyramidal neuron. We employ a full-truncation semi-implicit Euler scheme to control boundary violations and perform Monte Carlo parameter sweeps, time-step checks, burst-detection sensitivity analyses, and conductance perturbations. In subthreshold regimes, noise induces firing and produces an intermediate-noise minimum in the coefficient of variation, consistent with coherence resonance. In the deep subthreshold regime, the burst rate exhibits moderate Arrhenius-like scaling over a restricted low-noise interval, supporting an interpretation in terms of noise-activated escape. Near the numerically identified onset of sustained oscillations, temporal coherence is particularly sensitive to noise intensity. In the suprathreshold regime, strong multiplicative noise accelerates firing through premature stochastic exits from the hyperpolarized recovery branch, although this behavior does not follow classical Kramers scaling. Control simulations using clipped additive Gaussian noise produce qualitatively different dynamics, indicating that the state dependence and boundary behavior of the diffusion term materially affect the observed transitions. The main trends remain robust under the tested variations in numerical resolution, burst-detection threshold, slow timescale, and M-current conductance. These results characterize model-specific effects of boundary-constrained channel noise and motivate future comparisons with discrete-state ion-channel models.
Read the full abstract

Optimization and Control

BOOOM: Loss-Function-Agnostic Black-Box Optimization over Orthonormal Manifolds for Machine Learning and Statistical Inference

Optimization over the Stiefel manifold $\mathrm{St}(p,d)$, the set of $p \times d$ column-orthonormal matrices, is fundamental in statistics, machine learning, and scientific computing, yet remains challenging in the presence of non-convex, non-smooth, or black-box objectives. Existing methods largely rely on either convex relaxations or gradient-based Riemannian optimization, limiting applicability in derivative-free and highly multimodal settings. We propose BOOOM (Black-box Optimization Over Orthonormal Manifolds), a general-purpose framework for loss-function-agnostic optimization on $\mathrm{St}(p,d)$. The key idea is a global Givens rotation-based parametrization that maps the manifold to an unconstrained Euclidean angle space while preserving feasibility exactly. Building on this representation, BOOOM employs a structured, parallelizable, derivative-free search based on Recursive Modified Pattern Search, enabling systematic exploration through plane-wise rotations without requiring gradient information and facilitating escape from poor local optima. We establish a unified theoretical framework showing equivalence between angle-space and manifold optimization, transfer of stationarity, and global convergence in probability under mild conditions. Empirical results across diverse problems, including heterogeneous quadratic optimization, low-rank and sparse matrix decomposition, independent component analysis, and orthogonal joint diagonalization, among other widely studied settings, demonstrate strong performance relative to state-of-the-art methods, particularly in non-smooth and highly multimodal regimes. We further illustrate its practical utility through a novel supervised PCA formulation applied to metabolomics data in colorectal cancer.
Read the full abstract

Combinatorics

Graphs with $\{P_3,P_4,P_5\}$-factors in terms of size and spectral radius

Let $G$ be a connected graph of order $n$. A $\{P_3,P_4,P_5\}$-factor is a spanning subgraph $H$ of $G$ such that every component of $H$ is isomorphic to an element of $\{P_3,P_4,P_5\}$. In this paper, we establish a sufficient condition on the size of the graph $G$ with minimum degree $\delta$ to have a $\{P_3, P_4, P_5\}$-factor. Subsequently, we provide another sufficient condition on the adjacency spectral radius, ensuring that a connected graph $G$ with minimum degree $\delta$ contains a $\{P_3, P_4, P_5\}$-factor.
Read the full abstract

General Mathematics

Measles Resurgence in Bangladesh, 2026: A Situational Analysis for Urgent Public Health Response

Background Measles has resurged globally in the post-pandemic period as routine immunisation recovery remains below the two-dose threshold required to interrupt transmission. Bangladesh, previously nearing measles–rubella elimination, entered 2026 with widening coverage gaps, depleted vaccine stocks, and increasing numbers of missed children. We conducted a situation analysis to assess the scale, concentration, and programmatic implications of the outbreak. Methods We performed a rapid mixed-evidence review from 1–15 April 2026 using data from WHO, UNICEF, DGHS bulletins, PubMed/MEDLINE, ReliefWeb, SEARO updates, and Bangla-language media. Of 46 records screened, 19 were included. Analysis was based on aggregated, publicly available surveillance and programme data. Findings By 15 April 2026, Bangladesh reported 19{,}161 suspected cases, 2{,}973 confirmed cases, 166 suspected deaths, and 32 confirmed deaths across 58 districts since 15 March 2026. The outbreak was spatially concentrated: the top two divisions accounted for 56.5% of cases (HHI = 0.217). Children under five comprised 81% of cases, including 34% infants under nine months. Vaccination status showed 72% zero-dose and 16% partially vaccinated cases. Coverage declined from 88.6% to 86% for MR1 and from 89% to 80.7% for MR2 (2019–2024), leaving about 20 million children vulnerable. Interpretation The resurgence reflects accumulated immunity gaps rather than vaccine failure, driven by subnational inequities and programme disruption. Urgent priorities include targeted vaccination campaigns, restoration of vitamin A supplementation, strengthening paediatric care capacity, and integrating real-time surveillance into outbreak response.
Read the full abstract

All authors Md. Kamrujjaman, Faizunnesa Khondaker, Alvi Ahmed Sarker, Nuzhat Nuari Khan Rivu

Probability

Free compressions of R-diagonal random variables and the semigroup of Brown measures

We investigate the Brown measures of compressions of $R$-diagonal random variables, extending previous results to include unbounded cases. For random variables with finite variance, we demonstrate that the Brown measures of their compressions converge to the uniform distribution on the unit disc. In the case of infinite variance, we characterize the Brown measures that remain stable under the compression operation and explore their properties in detail.
Read the full abstract