Colleagues

Works on aspects of computational algebraic geometry, with particular interest in Moller-Stetter matrices and bivariate analogues of linearizations. We are working on making tools from algebraic geometry more numerically friendly.

Works on classical approximation theory, including the stability of perturbed quadrature rules, and on convergence rates and frequency bias in neural network training. Before Cornell, he studied mathematics at Vanderbilt.

Works on hyperbolic PDEs in the context of recovering underlying solution operators, and on infinite-dimensional linear algebra for improving data-driven techniques. Before Cornell, he studied at Columbia.

Works on continuous and randomized linear algebra, especially spectra-related properties of Green's functions and Koopman operators. DMD and Hutchinson-type estimators feature heavily in her work. Before Cornell, she studied engineering at UT Austin.

Works on rational neural networks, making them efficient and scaling them up to practical applications.

Works on Cholesky decomposition convergence at the continuous level and Mercer's expansion.

Worked with Paul Steen and me on spectral theories of inertial-capillary motions and numerical tools for spherical caps modeling droplets resting on surfaces. She is an expert on dynamical systems, queueing theory, and traffic-flow questions.

Now working in industry in Somerville, Massachusetts.

Worked on spectral element methods and finite element methods.

Now an Assistant Professor at the Institute for Mathematical and Computational Engineering (IMC) at the Pontificia Universidad Católica de Chile in Santiago.

Worked on matrix recovery, hierarchical off-diagonal low-rank matrices, and randomized linear algebra, with an eye toward theoretical questions in PDE learning. Before Cornell, she studied mathematics at Yale.

Now a Courant Instructor at New York University.

Worked on bifurcation diagrams, symmetry-exploiting numerical methods for PDEs, and active fluid simulation in spherical geometries. He spent time at Oxford and Cornell while collaborating on these problems.

Now at Imperial College London after postdoctoral work at Cambridge.

Worked on optimal-complexity spectral element methods, DG methods, and algebraic multigrid. He also helped build Wolfram Alpha and worked on multigrid methods at Disney.

Now a research scientist at the Flatiron Institute in New York.

Worked on continuous analogues of linear algebra algorithms, including Krylov-style methods for matrices coming from spectral discretizations of differential equations.

After work in augmented reality at Facebook, he is now an Assistant Professor at Princeton.

Worked on numerical methods for linear and nonlinear differential eigenproblems, exploiting ultraspherical spectral discretizations to build faster and more accurate eigensolvers and later algorithms for spectral measures.

Now on the faculty at RPI after serving as an Applied Math Instructor at MIT.

Worked on tensor formats, compression algorithms, and computing with tensors that have displacement structure, with broad interests spanning randomized numerical linear algebra, high-performance computing, and approximation theory.

Now a Research Scientist at Lawrence Berkeley National Laboratory.

Worked on numerical algorithms for Sylvester matrix equations with high-rank right-hand sides and created Diskfun. She was awarded NSF, diversity, and NASA fellowships during graduate school.

Now an Assistant Professor in Applied Mathematics at the University of Washington.

Worked on incorporating the poloidal-toroidal decomposition into numerical solvers for advection-dominated incompressible fluid simulations in polar and spherical geometries, co-supervised with Grady Wright.

Now a PhD student at Princeton after undergraduate study at MIT.

Worked on matrix equations, Zolotarev bounds, and data compression algorithms, with a strong emphasis on randomized linear algebra.

Now a PhD student in computer science at the University of Illinois Urbana-Champaign.

Worked on learning physical PDE models from experimental simulations, combining machine learning ideas with high-order PDE solvers, during a Fall 2019 internship at the Center for Applied Mathematics.

Worked on numerical algorithms for solving multivariate polynomial systems, with particular focus on Groebner-, border-, and H-basis approaches. He was also a top-200 finisher in the Putnam.

Afterwards a PhD student at MIT.

Worked on a new nonuniform fast Fourier transform based on low-rank approximation, with applications to spherical rotations, Chebyshev expansion evaluation, and univariate polynomial rootfinding.

Now a mathematics lecturer in Spain.

Worked on a spectral element method for meshes with skinny elements, exploiting ideas from the ultraspherical spectral method for singularly perturbed differential equations.

Now working in industry after undergraduate study at MIT.