Four New Faculty Members Represent a ‘New Wave’ of Computational Thinking in Math

Four new assistant professors in the Duke Department of Mathematics represent a great stride toward the future of math, Department Chair and Phillip Griffiths Professor of Mathematics Robert Bryant says.

“For all four of these new people, computation has fundamentally changed their approach to problems,” he says. “They are exploring areas that were simply out of reach in the past.”

The hires also reflect Math’s participation and support of Duke’s commitment to the Center for Computational Thinking. Duke’s vision is to be a leader in research based on computational technology and also to lead the training of the next generation of scholars, researchers and engineers.

“Our new faculty are working in really exciting areas,” Bryant says. "They represent a new wave in mathematics.” 

Teaching computation

Suzanne Crifo is an assistant professor of the practice who will focus on teaching computational mathematics.

Suzanne Crifo
Suzanne Crifo (John West/Trinity Communications)

Crifo says that she discovered the fun of teaching and working with students during her Ph.D. at North Carolina State University, when she was as “a very theoretical mathematician” studying representation theory of infinite dimensional Lie algebras. She thought the work was fun, “but I found myself drawn more to the teaching portion of graduate school because I could see how I was impacting students,” she says. “I sought out every professional development opportunity I had for teaching and applied to a lot of teaching-focused positions.”

She first came to Duke after graduating in 2019 not to teach, but to become a learning consultant in the Academic Resource Center.

“In the Academic Resource Center, I learned even more about good teaching that supports good learning,” Crifo says. “So I'm really excited now that I am a teaching-focused faculty to use everything I've gathered about the learners at Duke, and specifically the undergraduate learners.”

This Fall, Crifo will be teaching Laboratory Calculus and Functions I (Math 105L) and Python Programming in Mathematics (Math 260). In the Spring, her teaching will focus on Theory and Practice of Algorithms (Math 560), which includes students in the master's in data science program. She’ll also be available to help her faculty colleagues develop computational modules for their courses.

“I think the curiosity that I bring can be very helpful, especially for someone who's maybe been teaching a course for several semesters, to help them consider different angles,” she says.

Crifo earned a double BA in Math and Music as an undergraduate at the College of the Holy Cross in Worcester, Mass., though maybe the music came first. She is a classical soprano who counts Mozart and Gershwin as her favorites. A native of New Jersey, Crifo likes to exercise and hang out with her husband and dog.

Real-world randomness

Alex Dunlap is an assistant professor of Mathematics who applies stochastic partial differential equations to real-world problems.

Alex Dunlap
Alex Dunlap (John West/Trinity Communications)

A stochastic problem has an element of randomness to it, Dunlap explains. It might be modeling the heat flow in a room where there are a bunch of toasters going on and off at random. “The heat still follows all the rules of physics. But we're adding this sort of less predictable input,” says Dunlap.

“Oftentimes we use stochastics because we don't understand the physics at the smallest scale,” Dunlap says. “We might sort of understand the dynamics of a larger scale, but maybe we think something really is random, maybe it's something quantum with some kind of random effect. Or maybe it's just that something different and complicated is happening at the small scales, and we don't really understand it. So we say, ‘well, let's treat it as random,’ and that often turns out to be a good modeling choice.”

The stochastic partial differential equations he hopes to dive into at Duke will examine scale invariance in a problem by looking for areas of self-similarity. “I'm very interested in how the self-similarity allows us to actually compute some quantities exactly,” he says. “And I think that's pretty surprising.”  

Dunlap is excited to be joining Duke, where problems like the ones he tackles can be found in every direction.

“My work fits in pretty naturally with the work of several current faculty in the department,” Dunlap says. “I got the sense that (Duke has) a lot of interdepartmental collaborations and strong groups in things like Biology and Physics and Computer Science and Statistics, which are all areas I'd be excited to interact with.”

Dunlap completed his Ph.D. at Stanford and is joining Duke from a three-year postdoc at New York University. He is a rock climber and bicyclist who grew up in the Pacific northwest.

Quantum math

Di Fang is an assistant professor of Mathematics and a member of the Duke Quantum Center.

Di Fang
Di Fang

She is working on developing algorithms for a computer that is still being developed.  

"I believe that this is the perfect time to delve into numerical analysis of quantum algorithms,” Fang says. “Once we have access to these exceptionally advanced quantum computers, we’ll be ready to understand the precise applications they can be used for and how long it takes to deliver tangible and practical outcomes.

“Essentially, we are strategically laying the groundwork to harness the full potential of quantum computing for real-world problem-solving."

Fang’s expertise lies in applied and numerical analysis of partial differential equations with various applications. Her earlier work focused on classical algorithms for quantum problems and biological applications, but she caught the quantum computing bug at a Simons Institute program during her postdoctoral experience as a Morrey assistant professor at the University of California, Berkeley.

“Now, I'm most interested in quantum algorithms,” she says. “My greatest personal motivation is molecular dynamics from first principles, namely the many-body Schrodinger equation, which is a terribly high dimensional problem.”

Her recruiting visit to Duke’s campus included a tour of the Quantum Center in the Chesterfield Building in downtown Durham and meeting some of the experimentalists there. “All the places I was interested in had great quantum resources, but I feel like (Duke’s) level of interdisciplinary support is really, really precious,” she says.

Fang is also a talented classroom teacher, having won awards every year for her outstanding calculus teaching at the University of Wisconsin, where she earned her Ph.D.

“I try to use many examples to illustrate these rather abstract mathematical formulas to make them relevant to students’ daily life,” she says. Highly dimensional integrals, for example, are depicted by a package of cookies.

Fang earned her B.S. in Mathematics at Shanghai Jiao Tong University and has been supported by the National Science Foundation (NSF) and the Simons Foundation. She says she enjoys exercise, and now that she’s in Durham she just might resume the hip-hop dancing she did as an undergrad.

Making patterns

Fan Wei is an assistant professor of Mathematics seeking systematic ways to study very large combinatorial objects, like networks.

Fan Wei
Fan Wei (John West/Trinity Communications)

She joins Duke from Princeton University, where she has been an instructor.

“I study many kinds of combinatorial objects, and a network is one of them,” she says. “A social network, for example, is represented as a graph where people are nodes and friendships are connections between nodes.”

Math is needed to analyze a large network because it would require a ridiculous amount of printer paper to truly map out how the nearly 3 billion monthly users of Facebook connect with each other. Sampling and modeling helps, but even that isn’t much use, as the possible number of connections in such a network would be “larger than the number of atoms in the universe,” Wei says.

However, Wei’s research attempts to cut through the problem to find patterns, and analyzing which problems could or could not have a fast algorithm. A combinatorial object might also be permutations or rankings, or a sequence of words. “I also study other problems I find interesting, like game theory, for example,” she says.

Wei says she is looking forward to interactions with researchers in other fields at Duke. “When I was visiting, I gave a talk in the Computer Science seminar and people from different departments had joined and I got to talk with them,” she says. “I’m looking forward to talking to faculty and students from all different fields and trying to find common interests.”

Wei earned her Ph.D. at Stanford, a master’s from Cambridge and a bachelor’s degree from MIT. Her previous experience also includes postdocs at Princeton’s Institute for Advanced Study and Microsoft Research. She has been supported by the NSF and the Simons Foundation.

Wei enjoys watching Chinese movies and Korean and Japanese detective shows and is also a new mother. “So now I have two different new roles: as a mom and as a professor,” she says.