Those with interdisciplinary tastes should sample MALBEC

Doing math certainly takes brains, but does math tell us anything about the way our brains work? That’s one of the questions being explored this semester in an interdisciplinary seminar called MALBEC (Math, Algorithms, Learning, Brains, Engineering and Computing).

Sponsored by the departments of Mathematics and Computer Sciences with funding from the Morgridge Institute for Research and the Wisconsin Alumni Research Foundation, MALBEC features scientists who use computational approaches to understand the behavior, learning and perceptions of people and machines. On the one hand, this means exploring the relevance of mathematical models to the functioning of actual animal brains, says math professor Jordan Ellenberg. How well do these formal models correspond to our neurobiology, and what new insights do they offer?

But there’s another side as well, stresses Ellenberg, who is co-organizing MALBEC with math and biochemistry professor Julie Mitchell. It’s widely known that even the smartest computers struggle to accomplish what human brains do with ease, such as recognizing a face from a degraded image. The question then becomes: Can in-depth knowledge of animal brains help scientists design more nimble computers?

Addressing questions like these requires researchers to step outside the bounds of their fields, which is just what the series is designed to encourage.

“Rather than inviting our own faculty to speak, MALBEC is about looking outside to see whom we want to learn from,” says Mitchell. “It’s an opportunity to invite interesting people to come chat with us…and throw a party for them.”

In the future, the seminar may continue in different forms, add the organizers. Math and computer science will likely be mainstays, but the third partner is up for grabs: life sciences, social sciences, even art. The goal is to strengthen ties between mathematics and these less obviously computational fields.

“Physics and engineering are the traditional stomping grounds of applied mathematics,” says Ellenberg. “But if we’re open to it, I think there’s going to be a lot more room to stomp.”

The talks will be held in Van Vleck Hall and Computer Sciences. Visit this site for times and locations.

• Friday, April 17: Partha Niyogi, University of Chicago, studies pattern recognition and machine learning problems that arise from the computational study of human speech and language.
• Tuesday, April 21: Michael Coen, biostatistics and medical informatics, UW–Madison, studies self-supervised machine learning, based on biologically inspired models of learning in animals.
• Wednesday, May 6: Jerry Zhu, computer sciences, UW–Madison, studies the application of statistical machine learning algorithms to natural language processing, human computer interaction and cognitive psychology.