Welcome Christina Lee Yu
- New Faculty Year: 2018
Christina Lee Yu joined the faculty of the School of Operations Research and Information Engineering (ORIE) at Cornell University in July, 2018. The goal of Yu’s research is to design scalable statistical algorithms for processing social data based on principles from statistical inference.
“I am excited to be at Cornell,” says Yu. “I really enjoy that Cornell’s Operations Research department is very open-minded and interdisciplinary, with faculty working at the intersection of statistics, computer science, machine learning, economics, and applied math. Boundaries between disciplines are not drawn too closely here, which gives me flexibility to explore research in many different directions.”
Yu earned her undergraduate computer science degree from the California Institute of Technology. “Caltech was perfect for me,” says Yu. “It was a small school with a narrow focus on math, science and engineering, and I was given a lot of flexibility to choose courses across different departments. I was able to take a mix of CS, EE, and applied math courses that were tailored to my interests—it was the ideal mix to prepare me for research.”
During her junior year at Caltech, Yu worked with Professor Adam Wierman on a summer research project. “That research experience with Adam changed my perspective, “ says Yu. “It helped me see that maybe I could do research and it could be fun.” The research Yu did as an undergrad involved matching markets and social networks. “Wherever you look, there are complex networks,” says Yu. “For example, there are biological networks, transportation networks, communications networks, social networks. They are all intricate and patterned and seem unique, yet they share common properties. I applied to grad schools thinking I wanted to look deeply at why these networks seem to have so much in common.”
Yu received her M.S. and Ph.D. in electrical engineering and computer science from the Laboratory for Information and Decision Systems (LIDS) at the Massachusetts Institute of Technology (MIT). When she got to MIT her focus shifted to more of an engineering approach. “I started to ask, ‘How can we engineer optimal algorithms to perform inference over networks? Can we compute global properties of the network using local algorithms that can only see a small portion of the network at a time?’” says Yu. Yu’s work addressed fundamental questions like “What’s the minimum amount of data one needs to collect before being able to infer the underlying structure or pattern in a dataset?” and “What types of statistical assumptions on your data do you need to guarantee provable results for a recommendation algorithm?”
In deciding between academia and industry, Yu recalls “I once heard a faculty member say ‘While one can also produce great research in an industry lab, the goal of research in academia is not only the research itself, but in the process to train students. In some sense, this makes the primary product of academia the students.” That made a lot of sense to me, and resonated with my passions. I’m excited that academia offers a unique opportunity to make impact through mentoring and teaching students.”
Immediately before starting at Cornell, Yu held a postdoctoral researcher position at Microsoft Research New England. At Cornell, she has several lines of research underway. “One of the questions I am asking,” says Yu, “is ‘How do you deal with interdependence or correlations in your data? Since data is collected over time and recommendations are made sequentially—how can we account for the historical dependencies between the data we collect and the recommendations we make?’” Other areas that Yu is interested to explore include fairness in algorithms and designing recommendations for health care data.
Yu taught Statistical Principles in the Fall 2018 semester.
by Jackie Swift Nothing about Craig J. Fennie, Applied and Engineering Physics, is predictable or routine. “I stumble into things,” he says with a rueful smile. “There’s never a plan.” Yet Fennie’s... Read more about New Materials from the Atomic Scale Up