Mark Iwen
Associate Professor
 |
Contact information: |
I am an associate professor with a dual appointment in the Department of Mathematics and the Department of Computational Mathematics, Science and Engineering (CMSE) at Michigan State University. My research interests include computational harmonic analysis, mathematical data science, signal processing, and algorithms for the analysis of large and high dimensional data sets. For a somewhat buggy list of papers organized by citation count you can visit my Google Scholar profile.
In April of 2020 several colleagues and I founded the online One World Mathematics of INformation, Data, and Signals (MINDS) Seminar in response to the COVID-19 pandemic. You can find videos of the One World MINDS Seminar talks as well as links to many other great seminars on the seminar website above.
Research Topics and Talks
If you're interested in learning about some of the things I work on, you can watch any of the talks linked to in the list below on double speed (or even regular speed, to experience every awkward pause in real-time :).
- Want to be prepared for the next pandemic? You can start by watching this publicly accessible talk about group testing methods.
- Want to know how you can numerically solve a huge class of PDEs on 1000-dimensional domains with provable accuracy guarantees? I suggest using Sparse Spectral Methods.
- Want to know how you can numerically approximate functions of 1000-variables with provable accuracy guarantees? You can watch these videos of a short course I taught at the CRM in Montreal on Sublinear-Time Algorithms for Approximating Functions of Many Variables. In particular, this course will explain a lot about how the magic of sparse spectral methods (above) works.
- Want to see how to preserve all Euclidean distances to (and within) a given submanifold of high-dimensional Euclidean space in a near-optimally low-dimensional Euclidean embedding space? Surprisingly, it's not impossible if you use a terminal embedding (which is like a non-linear version of a Johnson-Lindenstrauss matrix on steroids).
Course Notes on Randomized Numerical Linear Algebra, Sublinear-Time Compressive Sensing, and Preliminaries (including Linear Algebra and Probability): A Mathematical Introduction to Fast and Memory Efficient Algorithms for Big Data. These notes are edited versions of notes scribed by several (former) students including Cullen A. Haselby, Craig Gross, and Eric Brodsky. They are aimed at advanced undergraduate and beginning graduate students. All comments, complaints, and suggestions are welcome.
Older versions are also available: Fall 2020 version 0.0