Metric Geometry
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
Tuesday 17 February
This note introduces the class of basic $r$-ball polyhedra in the $d$-dimensional Euclidean space $\mathbb{E}^{d}$ for $d>1$ and $r>0$. We investigate their face structure and, for given integers $0\leq i\leq d-1$, $n\geq d+1\geq 3$ determine the maximal number of $i$-dimensional faces among all basic $r$-ball polyhedra in $\mathbb{E}^{d}$ with $n$ facets. In addition, we establish that for $d>2$, every basic $r$-ball polyhedron is globally rigid with respect to its inner dihedral angles.
Information Theory
Phase-Resolved Molecular Timing Channels under Pulsatile Drift: Volterra Characterization and Corrected Inverse-Gaussian Approximation
Pulsatile drift makes molecular arrival times depend on the release phase. We study a one-dimensional absorbing timing channel with constant diffusivity and positive-mean sinusoidal drift. A moving-boundary formulation yields an exact Volterra characterization of the phase-conditioned first-hitting law, with regularity, uniqueness, properness, and moment properties. Within this framework, we integrate and assess the corrected inverse-Gaussian (C-IG) approximation introduced in our conference work. Comparisons with a numerically validated Volterra reference distinguish raw-kernel mass discrepancy from normalized shape errors and show that close cumulative agreement can coexist with density and peak-count discrepancies, particularly under transient backflow. Phase-family analysis further shows that a small between-phase contribution to delay variance can coexist with substantial differences between conditional cumulative distributions. We connect these conditional laws and protocol-induced mixtures to timing likelihoods and absorption probabilities within a prescribed reception window, clarifying the arrival-time structures preserved by C-IG and the phase information concealed by averaging.
Read the full abstract
Statistics Theory
Non-Stationary Covariance Functions for Spatial Data on Linear Networks
We introduce a novel class of non-stationary covariance functions for random fields on linear networks that allows both the variance and the correlation range of the random field to vary spatially. The proposed covariance functions are useful to model random fields with a spatial dependence that is locally isotropic with respect to the resistance metric, a distance that reflects the topology of the network. The framework admits explicit stochastic representations of the associated random fields and can be naturally extended to matrix-valued covariance functions for vector-valued random fields. We assess the statistical and computational performance of a weighted local likelihood estimator for the proposed models using synthetic data generated on the street network of the University of Chicago neighborhood.
Read the full abstract
Combinatorics
Point Count of the Top-dimensional Open Positroid Variety
In [GL24], Galashin and Lam discovered that when $k$ and $n$ are coprime, the proportion of subspaces in $\mathrm{Gr}(k,n)(\mathbb{F}_q)$ that lie in the top-dimensional open positroid variety $\Pi_{k,n}^\circ(\mathbb{F}_q)$ is $|(\mathbb{F}_q^\times)^n|/|\mathbb{F}_{q^n}^\times|$. In this paper, I recover this point count identity by relating the split torus action on $(\Pi_{k,n}^\circ)_{\mathbb{F}_q}$ and an anisotropic torus action on a $\mathbb{F}_q$ rational form of $\Pi_{k,n}^\circ$. The main step in the point count argument and the main technical result in this paper is that cyclic rotation acts trivially on the torus-equivariant cohomology of $\Pi_{k,n}^\circ$ when $k$ and $n$ are coprime.
Read the full abstract
Classical Analysis and ODEs
The general Brannan coefficient conjecture I: Watson-lemma approximations
The coefficients $A_n(\alpha,\beta,\omega)$ in the Maclaurin expansion $(1+\omega z)^{\alpha}(1-z)^{-\beta}= \sum_{n=0}^{\infty} A_n(\alpha,\beta,\omega)z^n$ are studied, where $\omega,z \in \mathbb{C}$ with $|z| < |\omega|=1$, and $\alpha,\beta \in (0,1]$. In 1973 Brannan conjectured that $|A_n(\alpha,\beta,\omega)|\le A_n(\alpha,\beta,1)$ for each positive odd integer $n$, and showed it is true for $n=3$. This has recently been proven for all odd integers $n\ge5$ by a number of authors in aggregate for the special case $\beta=1$. In this paper hypergeometric integral representations and Watson-type approximations are utilised, from which the general problem is reduced to numerically evaluating the minima of certain simple, explicit, slowly-varying functions over compact domains. From the positivity of these constants it is shown that the conjecture holds for $\alpha, \beta \in (0,1]$, $0 \le |\arg(\omega)| \le \pi-\phi_0$ and $n=5,7,9,\ldots$, where $\phi_0=0.061$.
Read the full abstract
Information Theory
On the Entropy of General Mixture Distributions
Mixture distributions are a workhorse model for multimodal data in information theory, signal processing, and machine learning. Yet even when each component density is simple, the differential entropy of the mixture is notoriously hard to compute because the mixture couples a logarithm with a sum. This paper develops a deterministic, closed-form toolkit for bounding and accurately approximating mixture entropy directly from component parameters. Our starting point is an information-theoretic channel viewpoint: the latent mixture label plays the role of an input, and the observation is the output. This viewpoint separates mixture entropy into an average within-component uncertainty plus an overlap term that quantifies how much the observation reveals about the hidden label. We then bound and approximate this overlap term using pairwise overlap integrals between component densities, yielding explicit expressions whenever these overlaps admit a closed form. A simple, family-dependent offset corrects the systematic bias of the Jensen overlap bound and is calibrated to be exact in the two limiting regimes of complete overlap and near-perfect separation. A final clipping step guarantees that the estimate always respects universal information-theoretic bounds. Closed-form specializations are provided for Gaussian, factorized Laplacian, uniform, and hybrid mixtures, and numerical experiments validate the resulting bounds and approximations across separation, dimension, number of components, and correlated covariances.
Read the full abstract
Functional Analysis
Clifford spectrum of three 2 by 2 matrices
We prove that the Clifford spectrum associated to three 2 by 2 matrices is nonempty. The structure of Clifford is described in terms "moving" level curves. We discuss some implication of a conjecture formulated for arbitrary size n by n of three matrices and give an example in the case of three self-adjoint operators in the infinite dimensional Hilbert space.
Read the full abstract
Differential Geometry
B.-Y. Chen's inequalities for Riemannian submersion and their applications
In this paper, we introduce B.-Y. Chen inequalities for Riemannian submersions between Riemannian manifolds. We derive these inequalities for vertical, horizontal, and mixed distributions, establishing relationships between intrinsic invariants and extrinsic invariants. We also investigate the corresponding equality cases. As applications, the results are obtained for submersions whose total space is a real, complex, generalized Sasakian space form. Several examples are provided to illustrate both equality and strict inequality cases.
Read the full abstract
Optimization and Control
Carleman Inequalities for the Heat Equation with Fourier Boundary Conditions: Applications to Null Controllability Problems
In this work, we establish a Carleman inequality for the heat equation with Fourier boundary conditions of the form $\partial_\nu y+by=f1_\gamma$, where the control acts on a small portion $\gamma$ of the boundary. We apply this inequality to address the null controllability problem with boundary control supported on this small region. An explicit solution to this problem is obtained via a system of coupled parabolic equations. Based on these results, we propose an iterative numerical method to solve the coupled system.
Read the full abstract
Number Theory
Szemerédi's Theorem Along Cantor Sets of Integers
Let $\mathcal C= \{k_1<k_2 < \cdots\}$ be Cantor set of integers, that is a set of integers with restricted digits modulo a base $b$, and suppose $0$ is one of the restricted digits. We show that $$ \liminf_{M-N\to\infty}\mathbb E _{n\in [N]- [M]} m(A\cap T^{-k_n} A \cap \cdots \cap T^{-\ell k_n} A )>0. $$ The averages above are uniform in the index $n$, and so the set of indices $n$ for which $m(A\cap T^{-k_n} A \cap \cdots \cap T^{-\ell k_n} A )>0$ is syndetic. By the Furstenberg correspondence principle, it follows that for every set of integers $A$ of positive upper Banach density there is a syndetic set $B$ of indices such that, for each $n \in B$, the set of $a$ with $a, a+k_n, \ldots, a+ \ell k_n \in A$ again has positive upper Banach density. A Cantor set contains an IP-set, so the existence of such progressions, and the positivity of the density above for a single index, already follow from the IP Szemerédi theorem of Furstenberg and Katznelson, and from the work of Kra and Shalom. The point of this paper is instead the frequency with which the step sizes $k_n$ may be found, measured in the natural enumeration of the Cantor set itself.
Read the full abstract
Statistics Theory
Bayes Risk for Goodness of Fit Tests
We develop a unified framework for goodness-of-fit (GOF) testing through the lens of Bayes risk. Classical GOF procedures are commonly calibrated either at fixed significance level (CLT scale) or through exponential error exponents (LDP scale). We establish that Bayes-risk optimal calibration operates on the moderate-deviation (MDP) scale, producing canonical $\sqrt{\log n}$ inflation of rejection thresholds and polynomially decaying Type I error. Our main contributions are: (i) we formalise the Rubin–Sethuraman program for KS-type statistics as a risk-calibration theorem with explicit regularity conditions on priors and empirical-process functionals; (ii) we develop the precise connection between Bayes-risk expansions and Sanov information asymptotics, showing how $\log n$-order truncations arise naturally when risk, rather than pure exponents, is the evaluation criterion; (iii) we provide detailed applications to location testing under Laplace families, shape testing via Bayes factors, and connections to Fisher information geometry. The organizing principle throughout is that sample size enters Bayes-optimal GOF cutoffs through the MDP scale, unifying KS-based and Sanov-based perspectives under a single risk criterion.
Read the full abstract
Dynamical Systems
Integer Cantor Sets: Arithmetic Combinatorial Properties
Cantor sets of integers have a rich set of arithmetic combinatorial properties. We consider classical Cantor sets, with a base and a fixed set of allowed digits. For such sets, we (a) give examples of such sets that satisfy the intersective property with power savings (b) characterize uniform distribution, (c) establish polynomial mean ergodic theorems and (d) study metric pair correlation of Cantor sets.
Read the full abstract
Mathematical Physics · Cross-listed from nlin.AO
On the efficiency of pairwise Hamiltonian control to desynchronize the higher-order Kuramoto model
Synchronization of coupled oscillators is observed in many natural and engineered systems and emerges due to the interactions within the system. It can be both beneficial, e.g., in power grids, and harmful, e.g., in epileptic seizures. In the latter case, efficient control methods to desynchronize the systems are crucial. Recent studies have shown that interactions are not always pairwise, but higher-order, i.e., many-body, and this greatly affects the dynamics. For instance, higher-order interactions increase the linear stability of synchronized states but simultaneously shrink their attraction basin, with potentially opposite effects on control methods. Here, we use a minimally invasive pairwise control based on Hamiltonian control theory, and investigate its efficiency on phase oscillators with higher-order interactions. We show that, if the initial phases are close to the synchronized state, higher-order interactions make desynchronization more difficult to achieve. Otherwise, a non-monotonic effect appears: intermediate strengths of higher-order interactions impede desynchronization while larger ones facilitate it. In all cases, the control can desynchronize the system with a sufficient number of controlled nodes and intensity.
Read the full abstract
Mathematical Physics · Cross-listed from hep-th
Compact Q-balls and Q-shells within a $CP^N$ Skyrme-Faddeev type model
While $CP^N$ models with analytic potentials are known to support finite-energy compact Q-ball and Q-shell solutions, their behavior in more complex Lagrangian frameworks remains a subject of active research. This work explores these non-topological structures within an extended Skyrme-Faddeev-type model that incorporates quartic derivative terms. In this context, harmonic time dependence and the presence of quartic terms constitute two independent stabilization mechanisms that allow the configurations to circumvent Derrick's scaling argument. We investigate the necessary conditions for the existence of these solutions and analyze the influence of quartic terms on the properties of the resulting compactons, specifically examining the $E(Q)$ relationship between energy and Noether charge. Our findings provide valuable insights into the stability and characteristics of compact boson stars within $CP^N$ models featuring higher-order derivative terms.
Read the full abstract
Dynamical Systems
Dynamics of perturbed elliptical billiard tables
Dynamical billiards consist of a particle on a two-dimensional table, bouncing elastically off the boundary. The state of the system is given by two numbers: one describing the location along the curve where the bounce occurs, and another describing the incoming angle of the trajectory before the bounce. Tracking these numbers over successive bounces defines a two-dimensional area preserving map, and iterating this map gives a dynamical system first studied by Birkhoff. Although there are powerful theoretical results showing that generic (strictly convex) billiards exhibit chaotic dynamics, it is nevertheless difficult (if not impossible) to decide when a given billiard table is generic and one will often resort to numerical computations. In this paper, we employ the parameterization method to compute high order Taylor expansions of the local stable/unstable manifolds attached to periodic orbits of billiard maps. Globalizing appropriate fundamental domains locates transverse intersections, providing insight into the existence and location of chaotic invariant sets. A key step in implementing the parameterization method is computing the composition of a given polynomial with the (implicitly defined) billiard map, and we show that this can be done efficiently using the discrete Fourier transform (DFT). The DFT requires evaluating the composition on a disk in the complex plane, so that we must first extend the billiard table/map to a complex domain. This problem is addressed from a computational perspective.
Read the full abstract
Numerical Analysis
A Patankar predictor-corrector approach for positivity-preserving time integration
Many natural processes, such as chemical reactions and wave dynamics, are modeled as production-destruction (PD) systems that obey positivity and linear conservation laws. Classical time integrators do not guarantee positivity and can produce negative or nonphysical numerical solutions. This paper presents a modular correction strategy that can be applied to stiffly implicit Runge-Kutta schemes, in particular stiffly accurate SDIRK methods. The strategy combines stage-wise clipping with a ratio-based scaling that enforces invariants and is guaranteed to yield non-negative, conservative solutions. We provide a theoretical analysis of the corrected schemes and characterize their worst-case order of accuracy relative to the underlying base method. Numerical experiments on stiff ODE systems (Robertson, MAPK, stratospheric chemistry) and a nonlinear PDE (the Korteweg-De Vries equation) demonstrate that the corrected SDIRK methods preserve positivity and invariants without significant loss of accuracy. Importantly, corrections applied only to the final stage are sufficient in practice, while applying them at all stages may distort dynamics in some cases. These results show that the proposed framework provides a simple and effective way to construct positivity-preserving integrators for stiff PD systems.
Read the full abstract
Monday 16 February
Optimization and Control
Optimization-based One-side Boundary Control of LWR Traffic Models
In this paper, we study the feasibility of a class of optimization-based boundary control of one-dimensional macroscopic traffic flow models, where stability and invariance are achieved by a single boundary control. We define the sets of controllers to stabilize the system to a desired state via Lyapunov functionals, and to ensure forward invariance of a desired subset via boundary control barrier functionals. The control input is then selected from the intersection of those sets via a convex optimization problem. We determine sufficient conditions to ensure the existence of an optimal boundary control problem achieving both stability and invariance for a generic traffic flux function. Simulation results showcase the behavior of the proposed optimization-based controller applied to conservation laws with several traffic flow functions.
Read the full abstract
Optimization and Control
A Homotopy Framework for Constrained Multiobjective Optimization
We develop a homotopy continuation framework for computing Karush–Kuhn–Tucker (KKT) points of constrained multiobjective optimization problems. The proposed homotopy map continuously deforms an easily solvable system into the KKT system associated with the original multiobjective problem, yielding a deterministic continuation path. Under the stated regularity and reference-point cone avoidance assumptions, we establish the existence of a smooth, bounded continuation path whose set of accumulation points as the homotopy parameter approaches zero is nonempty and consists of KKT points. The framework is implemented using two numerical continuation algorithms including an enhanced pseudo-arclength predictor–corrector strategy with adaptive step-size control and branch-restarting mechanisms to improve robustness for challenging nonconvex and degenerate problems. Numerical experiments demonstrate that the proposed method remains effective even when initialized from nonfeasible points, indicating practical robustness beyond the theoretical guarantees. Comparative studies evaluate the proposed framework against classical scalarization approaches and the evolutionary algorithms. Experiments on benchmark problems, disconnected and nonconvex multiobjective problems, analytically solvable three-objective problems, and a multiobjective IEEE 5-bus optimal power flow problem demonstrate the ability of the framework to compute KKT points in a variety of constrained settings. The results also highlight the effects of initialization and the limitations of continuation-based methods in recovering complete Pareto-front coverage. Overall, the proposed framework provides a deterministic and interpretable approach to computing Pareto-critical (KKT) points in constrained multiobjective optimization.
Read the full abstract
Differential Geometry
sup x inf Inequality on manifolds of dimension 5
We give an inequality of type sup x inf in dimension 5 for a Yamabe type equation.
Classical Analysis and ODEs
Perturbations of Cauchy differences
This paper investigates functional equations arising from perturbations of Cauchy differences. We study equations of the form \[ f(x+y)-f(x)-f(y)=B(x,y) \quad \text{or} \quad f(xy)-f(x)f(y) = B(x,y) \] where $B$ is a biadditive mapping, and also more general cases where the inhomogeneity depends on unknown functions \begin{align*} f(x+y)-f(x)-f(y)&= \alpha x y \\[2.5mm] f(x+y)-f(x)-f(y)&= \alpha (x y)\\[2.5mm] f(x+y)-f(x)-f(y)&= \alpha(x)\alpha(y). \end{align*} Our results extend previous work on the bilinearity of the Cauchy exponential difference by Alzer and Matkowski. We characterize solutions under various structural and regularity assumptions, including additive and exponential Cauchy differences, and show that solutions often reduce to additive functions, exponential polynomials, or combinations thereof. For Levi-Civita type equations, we provide explicit representations of solutions in terms of additive and exponential components. Furthermore, we determine conditions under which real-valued solutions exist and describe their forms. The paper concludes with open problems concerning generalized equations that cannot be solved by the methods presented here, suggesting directions for future research.
Read the full abstract
Combinatorics · Cross-listed from cs.SI
Overcoming Tight Constraints in Soft Happy Colouring
The Soft Happy Colouring (SHC) problem, a mathematical framework for identifying homophilic network structures, seeks to maximise the number of $\rho$-happy vertices, i.e., vertices with at least a proportion $\rho$ of neighbours that share the same colour. Because this NP-hard problem makes finding exact solutions intractable for large networks, probabilistic metaheuristics such as the Cross-Entropy (CE) method are suitable candidates. However, pure CE frequently suffers from stagnation of the probability distributions and non-convergence in high-dimensional spaces. To address this, we introduce CE+LS, synergising CE's adaptive learning with a fast, structure-aware local search (LS). By restricting the search exclusively to local optima, CE+LS learns from high-quality structural characteristics rather than raw random samples. We mathematically and empirically demonstrate that this search space reduction resolves CE's stagnation, yielding a convergent algorithm. Evaluating CE+LS across 28,000 Stochastic Block Model graphs, validated by non-parametric statistical testing, demonstrates that it consistently outperforms existing heuristic and memetic algorithms. Furthermore, benchmarking against the commercial exact solver, CPLEX, on real-world networks confirms that CE+LS identifies near-optimal configurations in a fraction of the required computational time for CPLEX. Crucially, CE+LS remains highly efficient even in the tight constraint regime, where comparative algorithms usually fail.
Read the full abstract
Mathematical Physics
Conductive Heat Flux Driven by a Pressure Gradient in Non-Maxwellian Reference States
Standard Navier–Stokes–Fourier theory and Maxwellian-based Grad 13-moment closures yield no independent pressure-gradient driving of the conductive heat flux in an isothermal, single-component gas in the hydrodynamic (small-Knudsen) regime. This absence is specific to the Maxwellian local-equilibrium weight. We show that when the closure is constructed about a generalized class of isotropic non-Maxwellian reference weights with finite fourth moment – characterized by a single shape parameter (a kurtosis-like moment ratio) that deforms the distribution continuously away from a Maxwellian – the small-Knudsen constitutive reduction retains a bulk pressure-gradient (barothermal) contribution to the conductive heat flux. This mechanism predicts pressure-driven conduction as a direct kinetic signature of non-Maxwellian equilibrium moment structure.
Read the full abstract
History and Overview of Mathematics
Rethinking how randomness is taught in K-12 statistics subjects
The effective teaching and learning of statistics persist as a challenge in K-12 education and has clear impacts in developing competence and confidence of students in entering STEM fields especially in today's digital age of data science. In this paper, it is proposed that the current pedagogy around teaching the concept of randomness, a core element of statistics, is incomplete. It is argued that expanding on the observer-relative nature of randomness will address commonly existing ambiguities on what is and what is not random, help in explaining more complex statistical concepts such as hypothesis testing, and impart greater awareness on the practical value of the practice of statistics in data science.
Read the full abstract
General Mathematics
Dynamics of four families of methods with the same weight function to solve nonlinear equations
We study the dynamics of four families of methods obtained with a weight function from a convex combination of Newton's method and a Newton-Halley type method on polynomials with two roots. We find the analytical expressions for the fixed and critical points. We study the stable and unstable behavior of the strange fixed points. Also, parameters spaces for identify methods with good behavior are presented. Then, several dynamic planes are presented to confirm the results obtained. Finally, some periodic orbits with period two for a selected method are presented.
Read the full abstract
General Mathematics
Analysis Of A Long Memory Circular Convolution Model
A stochastic model, the product of a circulant matrix and a random normal vector, is shown to produce an evolutive long memory time series with a power law spectral density. The distribution of the time series, a beta location scale family of distributions, provides a connection to the unit centered spherical distribution and directional statistics. The eigenanalysis of the deterministic circulant matrix is shown to provide estimates of the discrete Fourier spectral trend, the intrinsic dimension, the probability density shape parameter of the resulting time series, the condition number of the matrix and a principle component analysis. Examples of the R code, used as the constructive exploratory element of the work are given as constructive elements of the paper. The R code may be copied, pastedinto a R editor, and explored.
Read the full abstract
Mathematical Physics
Numerical Solution of the Bardeen-Cooper-Schrieffer Equation for Unconventional Superconductors
In this work, we consider the analytical properties and the efficient numerical solution of the Bardeen-Cooper-Schrieffer equation for unconventional superconductivity incorporating long-range power-law electron-electron interactions within a tight-binding model on a $d$-dimensional lattice. It is a nonlinear convolution equation for the complex matrix-valued superconducting gap under symmetry constraints imposed by the fermionic anticommutation rules. The long-range interaction enters in momentum space in the form of the now efficiently computable Epstein zeta function, which exhibits a power-law singularity at zero momentum. This needs to be accounted when evaluating the convolution. After a brief overview of some of the equation's analytical properties, we discuss its efficient numerical solution using a Galerkin method with B-splines. We present numerical results for a nodal superconductor on a two-dimensional square lattice.
Read the full abstract
Rings and Algebras
A short construction of the Lie algebra $G_2(K)$ over fields $K$ of characteristic $2$
The purpose of this paper is to give an explicit and elementary construction for the Lie algebras of type $G_2(K)$ of dimension 14, over the field K of characteristic 2. We say an elementary construction on the account that we use not more than little naive linear algebra notions.
Read the full abstract
Numerical Analysis
Quantum-Inspired Tensor Networks for Approximating PDE Flow Maps
We investigate quantum-inspired tensor networks (QTNs) for approximating flow maps of hydrodynamic partial differential equations (PDEs). Motivated by the effective low-rank structure that emerges after tensorization of discretized transport and diffusion dynamics, we encode PDE states as matrix product states (MPS) and represent the evolution operator as a structured low-rank matrix product operator (MPO) in tensor-train form (e.g., arising from finite-difference discretizations assembled in MPO form). The MPO is applied directly in MPS form, and rank growth is controlled via canonicalization and SVD-based truncation after each step. We provide theoretical context through standard matrix product properties, including exact MPS representability bounds, local optimality of SVD truncation, and a Lipschitz-type multi-step error propagation estimate. Experiments on one- and two-dimensional linear advection-diffusion and nonlinear viscous Burgers equations demonstrate accurate short-horizon prediction, favorable scaling in smooth diffusive regimes, and error growth in nonlinear multi-step predictions.
Read the full abstract
Number Theory
On the Lagarias Inequality and Superabundant Numbers
We study the Lagarias inequality, an elementary criterion equivalent to the Riemann Hypothesis. Using a continuous extension of the harmonic numbers, we show that the sequence $B_n=\frac{H_n+e^{H_n}\log(H_n)}{n}$ is strictly increasing for $n\ge 1$. As a consequence, if the Lagarias inequality has counterexamples, then the least counterexample must be a superabundant number; equivalently, it suffices to verify the inequality on the superabundant numbers.
Read the full abstract
Optimization and Control
Operating room planning with pooling downstream beds among specialties: A stochastic programming approach
In this paper, we study pooling downstream beds across specialties in a stochastic operating room planning problem. The main sources of uncertainty are stochastic surgical durations and patients' lengths of stay. We developed a two-stage stochastic programming model where in the first stage we decide on 1) the number of non-shared ICU and ward beds to be allocated to each specialty, and 2) the allocation of surgeries to operating rooms during the planning horizon. In the second stage, we decide on 1) the number of shared beds in ICU and wards to be allocated to different specialties on each day during the planning horizon, 2) the surge capacity required to satisfy downstream service to patients, and 3) the overtime incurred in operating rooms. The proposed model aims at minimizing the total cost including the patients' waiting cost, postponement cost, overtime and fixed cost of operating rooms, and the cost of downstream surge capacity. We have implemented the proposed stochastic programming model in a sample average approximation framework. To enhance the efficiency of sample average approximation, we have developed a specialized algorithm that quickly solves the second-stage model for any given first-stage solution for a large number of scenarios. We have carried out extensive computational experiments to evaluate the effectiveness of several pooling policies for downstream beds and also the efficiency of the proposed sample average approximation algorithm. Moreover, we have performed an extensive sensitivity analysis of cost and stochastic parameters. Our results demonstrated that a full-sharing policy among different specialties in the downstream units enhance the functionality of the system by up to 19.53%. Moreover, the results indicated that the solutions obtained by the proposed stochastic model outperform those from the corresponding deterministic problem by 17.43% on average.