First cohort of graduate students earn CSML certificates

Wednesday, Nov 13, 2019
by Sharon Adarlo

Anqi Wu

Anqi Wu

The Center for Statistics and Machine Learning (CSML) at Princeton University awarded certificates to its first cohort of graduate students this fall: two doctoral students and six finance master’s degree students, capping a year of milestones at the center.

“We are very proud of the achievement of our students,” said Peter J. Ramadge, the CSML director, the Gordon Y.S. Wu Professor of Engineering and professor of electrical engineering. “They constitute a mix of disciplines. That shows how data science can impact a variety of fields and its growing importance.”

The six finance students who graduated are the following: Xueyi (Erika) Hua, Luca Rona, Yingxue (Alicia) Zhou, Franck Stephane Ndzana Mvondo, Cyril Garcia and Melanie Bekx. The doctoral students are He Sun and Anqi Wu.

Sun, who graduated with a doctoral degree in mechanical and aerospace engineering, focused his research studies on optimizing the effectiveness of advanced space and large ground-based telescopes using machine learning techniques. Telescopes experience wavefront aberrations, a type of optical distortion that can blur an image. Sun uses machine learning techniques to remove wavefront aberrations in these large telescopes.

For his research, Sun said his CSML certificate coursework was very useful. 

“After I chose the machine learning certificate, I learned so many new ideas, and I have applied all of it into my research,” he said.

The largest telescope Sun has studied and worked on is WFIRST, Wide Field Infrared Survey Telescope, a NASA space telescope set to launch in the mid 2020s, he said. The telescope’s mission is to study dark matter and dark energy, find exoplanets and explore topics related to infrared astrophysics. 

Sun is continuing his studies as a postdoctoral researcher at Caltech’s department of computing and mathematical sciences.

Wu, who earned her doctoral degree in neuroscience, studied how to use statistical machine learning methods to analyze neural data. 

“We want to understand what the neurons are doing, and how neural activity correlates to behavior and external stimuli,” she said.

Machine learning techniques are essential to studying neural data because the information researchers receive straight from the brain is noisy and highly dimensional. You can’t just look at this data in a straightforward manner, Wu said. You need machine learning tools to find patterns.

“It’s a very challenging and interesting problem,” Wu said about why she got into her field of research.

And her CSML certificate proved to be valuable in completing her studies, said Wu, who is now working as a postdoctoral researcher at Columbia University.

Zhou found her experience at CSML to be fulfilling as well. She is now working as a data analyst in the market intelligence team of Point 72 Asset Management, a hedge fund firm.

“I had a wonderful experience at CSML,” said Zhou. “The staff, especially Professor Ramadge and Susan Johansen (CSML academic program coordinator), were extremely helpful in the progress of completing the certificate.”

Zhou’s research focus on campus was building investment strategies using quantitative methods in machine learning and econometrics. 

He Sun

He Sun

Bekx said she enjoyed the intellectual rigor of her time at CSML and how it expanded her worldview on what was possible for data science.

“The graduate certificate from CSML was a great experience to think more broadly about my research. Not only is it eye opening to hear about the research that is conducted in other fields than my own, it is also an opportunity to learn about different ways to tackle a research problem,” she said. 

Bekx is currently a strategist at Goldman Sachs and plans on continuing working in financial services.

“Trying to find patterns in data is exciting, and I'd love to continue applying the statistics and machine learning tools I learned to explain what happens in financial markets and continue modelizing some of the market's dynamics,” she said.

The center awards the graduate certificate in statistics and machine learning as a complement to students’ departmental graduate studies. Students must fulfill three requirements to earn the certificate: complete appropriate course work, engage in research involving statistics or machine learning, and participate in SML 510: Graduate Research Seminar. This course requires graduate students to present seminars on various research topics, share and learn from each other’s scholarly pursuits, give and receive feedback, and practice for the preparation and delivery of public presentations. Select faculty and external visitors also attend to give seminars on data-driven research. 

Zhou said she found the graduate seminar to be a great opportunity.

“I learned a lot from the weekly seminars, where students gave one-hour long presentations on how they apply data science (methods) in their own research. Since there are people from my program (Masters in Finance) and outside of my program, I walked away with a deeper understanding of how data science can be applied in finance, and how data science is applied in other fields, such as neuroscience and politics, which I did not have much idea about before,” she said.

The awarding of graduate certificates capped several months of growth and expansion at the center this year. In February, the University announced that the center is overseeing part of a wide-ranging and ambitious initiative that is set to spread and deepen the use of data science across campus. The center also celebrated its annual undergraduate poster session in May, which drew the participation of 62 students, the largest in its history.