Associate Professor Jessi Cisewski-Kehe has received a National Science Foundation (NSF) CAREER Award, the NSF’s most prestigious award for early-career faculty members. Her research is focused on developing innovative methods using Topological Data Analysis (TDA), which enables researchers to look for patterns in noisy datasets by scrutinizing the shape of the data in space—and looking for holes.
As a leading scholar in the field of astrostatistics, Cisewski-Kehe uses advanced statistical techniques to solve astronomical problems, such as uncovering the existence of Earth-like exoplanets. With the help of the CAREER Award, she will seek to improve existing TDA methodologies to analyze larger astronomical datasets—improvements that could also aid other fields.
Cisewski-Kehe discussed what the NSF CAREER Award means to her and the research it will allow her to take on, how statisticians and astronomers can learn from one another, and her guidance for students and younger statistics and data science researchers.
What does receiving an NSF CAREER Award mean to you?
I was very excited when I received the notification that my NSF CAREER proposal will be funded. It will allow me to pursue research and educational projects that I find interesting, worthwhile, and fun!
Your CAREER Award is entitled “Statistical Advances in Topological Data Analysis with Applications to Astronomy.” Could you explain the goals of your work in this area?
Topological data analysis (TDA) is a framework that seeks to identify and use shape information for visualization, inference, or prediction. My work focuses on persistent homology, a method that tracks different dimensional holes in a dataset, in which zero-dimensional holes are connected components, while one-dimensional holes are closed loops, and two-dimensional holes are voids, like the inside of a soccer ball.
With the CAREER award, I am seeking to extend persistent homology to larger datasets, such as cosmological simulation data that can include billions of particles, to build sound representations of detected holes in data that can aid with visualization and scientific discovery, and to establish statistical techniques that complement TDA methods for realistic time-series data.
On the education side, I plan to build resources to help statisticians and data scientists engage in astronomy research, involve undergraduate students in research, and develop a board game that encourages and cultivates statistical thinking in society.
How do you see statistics and data science more broadly contributing to our understanding of the cosmos?
Data from astronomy and cosmology can range in size from small to massive. For instance, the Ten-Year Legacy Survey of Space and Time will produce tens of terabytes of data each day, including trillions of observations of billions of galaxies. Astronomy data can take the form of point clouds, functions, images, manifolds, time series, spatial point processes, and more, so there is a lot of room for statisticians and data scientists to make a variety of significant contributions.
While many astronomers are capable of implementing existing statistical or machine learning methods for their analysis, there are often unique complexities that require novel methodology or, at least, sophisticated adjustments to accommodate them. While I think statisticians and data scientists can be helpful to astronomers to solve some of their pressing challenges, these challenges also give us new motivation for statistical methodology and theory.
What excites you most about bringing statistical expertise to interdisciplinary research?
I love that in statistics and data science we can develop a deep understanding of other disciplines through collaboration. Since my childhood, I have loved outer space. I remember mapping the northern hemisphere star constellations on my wall with glow-in-the-dark stars and doing a science fair project on black holes (I still have a copy of a letter I wrote to NASA asking them to mail me data!)
The fact that I get to work with astronomers and learn more about the Universe is so enjoyable! I love that I can potentially contribute to solving some of their data analysis challenges. It is a substantial investment to learn enough about a new field to be able to converse with the scientists, but I find it rewarding, exciting, and motivating.
As an associate professor and now an NSF CAREER award recipient, what guidance would you give to students or younger researchers who are interested in pursuing a career in statistics or data science?
An education in statistics and data science can open doors for a wide variety of careers in academia, government, or industry. If you have an interest in working with data, then I would recommend pursuing it. If you have a number of career directions you are interested in and cannot decide what to pursue, a nice feature of statistics and data science is that you can work in a broad range of fields using the skills you gain in our disciplines.
Regardless of your career choice, it can be helpful to find mentors that you trust who will encourage and support you, but also who will give you critical feedback. Two final tips that I try to keep in mind are (1) it is okay to say no to good opportunities if they do not contribute to the vision you have for your career or life, and (2) don’t be too good at things you don’t want to do.
To learn more about Associate Professor Jessi Cisewski-Kehe’s research, visit her website.