Oden Institute is the "Silicon Valley for Computational Math, Science and Engineering" – Profile Peng Chen

Peng Chen joined the Oden Institute as a Research Associate in 2015 and was promoted to a Research Scientist in February 2022. His research expertise falls in the areas of numerical analysis, scientific computing and machine learning. Chen obtained his bachelor’s degree in Mathematics and Applied Mathematics from Xi'an Jiaotong University in China, after which he moved to Switzerland to complete both his master’s degree in Mathematical Sciences and PhD in Computational Mathematics at École polytechnique fédérale de Lausanne (EPFL). He then spent a year conducting research as a postdoc at ETH Zurich before coming to the Institute in 2015.

Chen first discovered his interest in math while in elementary school. “I read a story about a little guy teleporting all around the world and through time to visit famous mathematicians—Zu Chongzhi, Aristotle, Archimedes, Euler, Bayes, Newton,” he said. “The little guy was fascinated by all kinds of interesting math problems and solutions from his conversations with the mathematicians, and so was I. Since then, I enjoyed math and continued to study it in college.”

During this time a research experience with a physics professor ignited his interest in computational math and its application in science. “The professor suggested figuring out why many metallic balls inside a box of oil under high electric field always form a fractal tree structure,” he said. “I dug into several papers that tried to explain this phenomenon, one using a set of about twenty equations from first principles. It was amazing to me,” he said. “Though I was not able to solve these equations, we proposed a computational method to characterize the fractal shape that led to a journal paper.”

Due to the interdisciplinary nature of computational science and engineering, teamwork and collaboration are integral to a successful project. This quality was appealing to Chen, who led a team of three in a Mathematical Contest in Modeling. The contest was in 2007 in China, and the team consisted of Chen, Chenglong Zhang (a former PhD student of Irene M. Gamba at Oden Institute) and Simeng Gao. After training for about 8 weeks, the team came up with a prediction for the Chinese population in the 21st century for the contest.

“We kept searching, reading, thinking, discussing, coding and writing. With full power propelling for 72 hours, we finished a paper using both computational differential equations and statistical Monte Carlo simulations,” Chen said. “We were selected as one of the top ten teams from several thousands, but more importantly, we developed a close friendship."

Then, during his graduate study and research, he was excited to see the pivotal role that computational math plays in many areas. For example, computational methods in medicine can help to understand blood flow and treating diseases in human cardiovascular systems through modeling, simulation, inference and optimization. This, in turn, can help doctors with the appropriate treatment, such as by designing the optimal shape of a bypass graft to restore normal blood flow in a partially blocked coronary artery.

“One specific project I worked on was simulation-based (using a stochastic fluid-structure interaction model) uncertainty quantification and sensitivity analysis of a relatively complete human arterial network, which helped to understand potential physiological and pathological implications,” said Chen, adding, “The beauty, power and practical use of computational math has inspired me to pursue a career in this field.”

Chen first discovered the Oden Institute when Björn Engquist visited Chen’s undergraduate university in China. Then, while he was a PhD student at EPFL, Chen often came across UT Austin and the Oden Institute in publications, conferences and SIAM News. Two of his colleagues, Fabio Nobile and Luca Dede’, had also come to the Institute to complete their postdoctoral research under Ivo Babuška and Thomas J.R. Hughes, respectively.

Chen said, “A lot of ideas and methods have been originated and developed from the Oden Institute. It was and has been the Silicon Valley to me for computational math, science and engineering. All these factors drew me to join the Oden institute.”

 

Chen works in Omar Ghattas’ Center for Computational Geosciences and Optimization (CCGO). He had attended a plenary talk given by Ghattas and was incredibly impressed by the powerful and scalable algorithms that enabled the computation to solve large scale Bayesian inverse problems in supercomputers (“the largest I’d ever seen,” said Chen) with application to the understanding of Antarctic ice sheet flow.

Chen is currently working on several projects including computational problems of Bayesian inference, stochastic optimization, optimal experimental design, and surrogate models. His research has a wide range of real-world applications, such as stellarator optimization for plasma fusion, optimal design of self-assembly materials, robust design of metamaterials, optimal sensor placement for tsunami early warning, inference of gravitational wave and optimal mitigation of infectious diseases.

“The specific goals are developing scalable computational algorithms to solve these problems constrained by large-scale complex physical systems under high-dimensional uncertainty,” Chen explained. “One specific project is a Simons Collaboration on stochastic optimization for the design of the shape of stellarator devices, which facilitate magnetic confinement of plasma fusion in order to harness nuclear fusion power.” In order to do this research, Chen often uses Frontera, one of the most advanced supercomputers at Texas Advanced Computing Center (TACC), to run computations.

Outside of work and academia, Chen enjoys playing badminton, hiking, and telling math stories to his son and daughter.