Matthew Williams Email & Phone Number

Honolulu, Hi

Princeton, New Jersey

Currently developing algorithms that blend techniques from machine learning and data mining with Koopman spectral analysis. These techniques approximate the leading Koopman eigenvalues, eigenfunctions, and modes, which are the nonlinear generalization of an eigenvector/eigenvalue decomposition for a linear system. Also exploring applications of these algorithms such as data fusion/merging, transport barrier identification, and coarse variable selection.

Bellevue, Washington

Applied multi-scale computing techniques, such as the Equation Free (EF) Framework, to problems in epidemiology related to the spread of Polio in the near eradication regime. Also developed the GPU based “detailed” code required for EF simulation using a combination of Python and OpenCL. The application of EF techniques reduced the number of detailed simulations required by between a factor of two and four, and had a minimal impact on accuracy.

Phd, Applied Mathematics

Thesis title: Exploiting Low Dimensionality in Nonlinear Optics and Other Physical Systems Adviser: J. Nathan Kutz Developed a data-driven.

Master Of Science (Ms), Applied Mathematics

Bachelor Of Science (Bs), Physics