Applied Math Seminar
The Applied Math Seminar hosts a wide range of talks in fields such as applied analysis, mathematical biology, fluid dynamics and electromagnetics, numerical computation, etc.
The seminar usually meets at 2:30pm on Fridays in room 1302 of Warren Weaver Hall.
Please email oneil@nyu.edu with suggestions for speakers. If you would like to be added to the mailing list, please send an email to cims-ams+subscribe@nyu.edu from the address at which you wish to receive announcements.
Seminar Organizer(s): Peter Nekrasov, Mike O'Neil, and Hanwen Zhang
Upcoming Events
-
Friday, December 4, 20262:30PM, Warren Weaver Hall 1302
TBD
Marc Gilles, Princeton University
Past Events
-
Friday, November 20, 20262:30PM, Warren Weaver Hall 1302
Fast and accurate electrostatics with edges and corners: Applications to quantum materials
Xuanzhao Gao, CCM, Flatiron InstituteSynopsis:
Solving for electrostatic fields in the presence of piecewise-constant media such as dielectrics and metals is a common engineering need. However, singular fields and surface charge densities at edges and material junctions make these problems challenging. In this talk, I will discuss a new boundary integral solver for these problems and its applications to quantum materials. Modeling modern quantum materials experiments can require solving thousands of boundary-value problems with charge sources derived from atomic orbitals. We reformulate the electrostatic transmission problem as a second-kind boundary integral equation and solve it efficiently using adaptive Nyström discretization, specialized near-field quadrature, fast multipole methods, and nonuniform fast Fourier transforms. We apply the method to a realistic device with graphene on a dielectric junction. However, the efficient and accurate handling of edges and corners in 3D BIE methods has been a persistent obstacle. I present a new quadrature method for the polyhedral case, based on known edge singularity families, and show that it can reach close to machine accuracies with surprisingly few unknowns. For example, we compute the capacitance of the PEC cube to 15 digits in about 30 seconds, the highest accuracy known to date. The method is expected to have a wider impact on BIE applications in viscous flow and wave scattering problems.
Joint work with Alex Barnett, Jason Kaye, and Malte Rösner.
-
Thursday, November 19, 202612:30PM, Warren Weaver Hall 1314
Leaky coins: how holes steer and tame the free-fall of disks, from translation–rotation coupling to path instabilities
Francois Gallaire, LFMI, EPFL, Lausanne, SwitzerlandSynopsis:
Freely falling disks follow a rich variety of paths, from steady descent to fluttering and tumbling. We examine how porosity reshapes these regimes. The porous disk is modelled by a homogenized interface whose permeability and slip coefficients are computed once and for all from unit-cell Stokes problems. This links the pore microstructure to the macroscopic flow without resolving the full-scale geometry.In the inertialess limit, a non-uniform porosity distribution breaks a planar symmetry and induces translation–rotation coupling. By tuning the pore size distribution, the same macroscopic disk can be designed to settle, drift or flutter.
At finite Reynolds numbers, permeability not only shifts the onset of the coupled fluid–solid instabilities of the steady vertical path but also selects the emerging mode. Beyond a critical permeability, all modes are stable and the disk falls steadily even at large Reynolds numbers. This stabilising effect is supported by experiments on perforated plates.
In collaboration with G.A. Zampogna, P.G. Ledda, E. Boujo, S. Camarri, J. Sánchez Rodríguez, G. Vagnoli and M. Froment.
Notes:
SPECIAL DAY/TIME !!!
-
Friday, November 13, 20262:30PM, Warren Weaver Hall 1302
Math and Geometry at McNeel
Sam Potter, Robert McNeel & AssociatesSynopsis:
Robert McNeel & Associates is a small software company, based in Seattle, with employees worldwide. Its main product is Rhinoceros 3D, used in architecture, engineering, industrial design, and many other smaller niche industries. I will discuss McNeel and Rhino broadly, and then talk in more detail about the work I do as one of a small team of CAD kernel engineers working on Rhino's propriertary, 25 year old CAD kernel. The bulk of this work involves an effort to modernize and improve Rhino's kernel (used by hundreds of thousands of people daily) to ensure that it helps Rhino stay relevant for the next 25 years. A key part of our approach is a shift towards verified numerical methods, chief among which is interval arithmetic. Simultaneously, this involves introducing an extensive and systematic approach to testing. As a case study, I will explain my current approach to computing verified Hausdorff distances between curves as a part of regression testing a newly rewritten intersection kernel (this latter rewrite being my main focus at McNeel). Time permitting, I will also showcase a few unrelated side projects taking place at McNeel. This talk is meant to give a
high-level overview of the kind of work taking place on the kernel team at McNeel. -
Monday, November 2, 20263:45PM, Warren Weaver Hall 1302
How do simple rotations affect the implicit bias of Adam?
Rebecca Willett, University of ChicagoSynopsis:
Adaptive gradient methods such as Adam and Adagrad are widely used in machine learning, yet their effect on the generalization of learned models – relative to methods like gradient descent – remains poorly understood. Prior work on binary classification suggests that Adam exhibits a “richness bias,” which can help it learn nonlinear decision boundaries closer to the Bayes-optimal decision boundary relative to gradient descent. However, the coordinate-wise preconditioning scheme employed by Adam renders the overall method sensitive to orthogonal transformations of feature space. We show that this sensitivity can manifest as a reversal of Adam’s competitive advantage: even small rotations of the underlying data distribution can make Adam forfeit its richness bias and converge to a linear decision boundary that is farther from the Bayes-optimal decision boundary than the one learned by gradient descent. To alleviate this issue, we show that a recently proposed reparameterization method – which applies an orthogonal transformation to the optimization objective – endows any first-order method with equivariance to data rotations, and we empirically demonstrate its ability to restore Adam’s bias towards rich decision boundaries. This is joint work with Adela DePavia and Vasileios Charisopoulos.
Notes:
NOTE SPECIAL TIME FOR AMS
-
Friday, September 25, 202610AM, Warren Weaver Hall 1302
Stability and Bifurcations in a Free Boundary PDE: Models of Cell Motility
Leonid Berlyand, Penn StateSynopsis:
We begin with a brief overview of the rapidly developing research area of active matter (a.k.a. active materials). These materials are intrinsically out of equilibrium resulting in novel physical properties whose modeling requires the development of new mathematical tools. We focus on studying the onset of motion of a living cell (e.g., a keratocyte) driven by myosin contraction. We introduce a minimal two-dimensional free-boundary PDE model that captures the evolution of the cell shape and nonlinear diffusion of myosin.
We first consider a linear diffusion model with two sources of nonlinearity: Keller-Segel cross-diffusion term and the free boundary that models moving/deformable cell membranes. We establish asymptotic linear stability and derive an explicit formula for the stability-determining eigenvalue. Next, we consider the effect of nonlinear myosin diffusion, which results in the change of the bifurcation type from super- to subcritical, and obtain an asymptotic representation of the bifurcation curve. This allows us to derive an explicit formula for the curvature at the bifurcation point that controls the bifurcation type. In the most recent work in progress with the Heidelberg biophysics group, we study the relation between various types of nonlinear diffusion and bistability. Finally, we discuss novel mathematical features of this free boundary model with a focus on non-self-adjointness, which plays a key role in the stability analysis. We conclude by presenting an example when the presence of a spectral gap does not guarantee stability and contrast this example with non linear stability in our minimal model.
Joint works with A. Safsten & V. Rybalko (Transactions of AMS 2023, and Phys. Rev. E 2022), with O. Krupchytskyi &T. Laux (Journal of nonlinear Scinence 2026), and with A. Safsten & L. Truskinovsky (ARMA 2026), with A. Safsten (2026-in progress). This work has been supported by NSF grants DMS-2404546, DMS-2005262, and DMS-2404546.
Notes:
NOTE SPECIAL TIME