Home

Welcome to my research page. I am a PhD student in applied mathematics at Purdue University. Previously, I was also affiliated with UIUC, UC Santa Barbara, and UCLA, where I was very fortunate to have had mentors of Xiaohui Chen, Paul Atzberger, and Chris Anderson.

I study pure and applied mathematical foundations of artificial intelligence. My research interests can best be categorized as scientific machine learning, geometric deep learning, and deep learning for differential equations (ODEs, PDEs, SDEs), although my research interests are broad and intersect other areas of mathematics but in ML.

The goal of my research has more recently become a means to address questions of artificial intelligence as a new technology, such as for example can we compute traditionally non-computable problems in high energy physics using methodologies from scientific machine learning? Can we detect cancer faster, more efficiently, and at scale using mathematics? Can we improve the Transformer, not with computing power, but by reinventing the mathematics of the attention mechanism? My current interest is quantum chemistry.

Please do feel free to add me on LinkedIn if you think you know me from somewhere, or please do feel free to shoot me an email.

At UIUC, I was affiliated with the Grainger College of Engineering through DIGIMAT, a research group focused on the interface of data science, statistical physics, and materials science. DIGIMAT is co-hosted by the National Center for Supercomputing Applications and Materials Research Lab.

Previously, I was a member of Atzberger Research Group, a group focused on statistical physics, numerical analysis, scientific computation, and machine learning.

One can reach me by email. My university email is agracyk at purdue dot edu, but my permanent email is andrewgracyk at gmail dot com. Also, I receive a lot of mass email, so it is possible (but unlikely) unfamiliar email is overlooked. LinkedIn is a great place to contact me.

.

Sunset on a glacial fjord along the Norwegian sea