News & Events


Featured Researcher: Hyungsuk Tak

Posted on September 13, 2019

Get to know Dr. Hyungsuk Tak, an ICS co-hire with the Eberly College of Science who joined Penn State in fall 2019 in the departments of Statistics and Astronomy and Astrophysics.

Dr. Tak is an astro-statistician—he uses complex data analytics methods to understand the Hubble constant, which measures the expansion rate of the universe.

How did you get into this research field?

My career as an astro-statistician began just because of free Chinese food. During my PhD days, the course “astro-statistics” provided weekly Chinese catering outside its classroom. One day I was picking up (stealing) remaining foods after the class starts because I did not take that course, and by chance I made an eye contact with the instructor. I felt guilty, so I entered the classroom. After the class ended, the instructor asked me whether I was interested in an astro-statistics project. When I told him that I had no idea about astronomy, he said “It is all about the data!” That was how and why I became an astro-statistician.

What do you hope to accomplish with your research? 

I want to develop and disseminate statistically rigorous and practically motivated data analytic tools for astronomy/astrophysics. My career goal is to estimate the Hubble constant, the current day expansion rate of the Universe, using my own methods.

How does supercomputing enable your research?

There are possibly many ways to infer the Hubble constant, but the method that I use is quite data-driven. Two types of data are produced from a strong gravitational lensing effect; multiple time series data and high-resolution imaging data. Analyzing both types of the data simultaneously in an iterative algorithm such as Markov chain Monte Carlo is challenging and computationally expensive, and thus I have benefitted from the high-performance ICS cluster.

What is your academic background?

I got PhD in statistics at Harvard in 2016, did post-doc at one of NSF-funded research institutions called SAMSI (Statistical and Applied Mathematical Sciences Institute) for two years, worked at Notre Dame as an assistant professor for a year, and finally joined Penn State via ICS co-hire this year.

What are the big problems you hope your research solves — and/or the big opportunities you hope your research seizes?

Again, my career goal is to estimate the Hubble constant by developing tools to handle various challenges. The most challenging issue that I anticipate is the advent of the big data astronomy because the Large Synoptic Survey Telescope (LSST) will produce 15 terabytes of imaging data every night from 2022 to 2032. Developing scalable methods is indispensable because existing data analytic tools may not work on the entire 5.5-petabyte data of the LSST. This challenge is definitely the most promising opportunity, and I will focus on developing a tool to infer the Hubble constant that can adapt to such big astronomical data.

What’s your favorite sound?

I like hearing the sound of brook water flowing, calming and relaxing.

If you were a computer part, what part would you be, and why? 

RAM because I often forget various things about past projects when I focus on new things.

What’s your advice for would-be scientists?

For a data scientist, it may be important to know what you are using in detail, instead of blindly running a publicly available package code. This is because science is built on assumptions, and the methodology is basically a set of assumptions possibly with various limitations. Doing science without fully mastering a data analytic method that you are using can be tempting, but it will make your science vulnerable in the end.

Favorite hobbies/pastimes that have nothing to do with your professional work?

I like reading books and running.

What is something that people are surprised to discover about you?

I was good at shooting a gun and handling hand-grenade, though not at the level of killing machine, as most Korean males are trained during the mandatory military service. I talked about this during an introduction session of my first year PhD, and my PhD days were very peaceful because no one tried stressing me out 🙂


Related Posts