Meet ML@GT: Minshuo Chen Wishes He had Started Studying Machine Learning Sooner

The Machine Learning Center at Georgia Tech (ML@GT) is home to nearly 100 talented students from across campus, representing all six of Georgia Tech’s colleges and the Georgia Tech Research Institute (GTRI).

These students have diverse backgrounds and a wide variety of interests both inside and outside of the classroom. Today, we’d like you to meet Minshuo Chen, a fourth-year machine learning Ph.D. student who wants to learn to recover a 3×3 Rubik’s cube in less than six seconds and is drawn to how math can help solve real-world problems.

Advisor: Tuo Zhao, assistant professor in the H. Milton Stewart School of Industrial Systems and Engineering and Wenjing Liao, assistant professor in the School of Mathematics

Hometown: Hangzhou, Zhejiang Province, China

Major/year: Fourth-year Ph.D. student in Machine Learning

Previous degrees earned and from what institutions: Bachelor’s degree from Zhejiang University and master’s degree from UCLA, both in electrical engineering.

Tell us about your research interests. Where might people be impacted by them in everyday life?

My research focuses on developing principled methodologies and theoretical foundations of machine learning and statistics. I am particularly interested in statistical, approximation and stochastic optimization theories of deep learning.

What drew you to wanting to research these areas?

Deep learning — sometimes coined as alchemy — stimulates exciting controversies in interdisciplinary areas like statistics, optimization, and applied mathematics. Yet, the theoretical foundations of deep learning are curiously under-explored. I find great passion and interest in pursuing the principles and backbone mechanisms within and behind deep learning. It looks like crossing the river by feeling the stones; the process is fascinating though, at least, when you feel a stone bolstering you.

Are you a part of any labs? If so, tell us about them and why you chose to work in those labs.

I am one of the oldest members of the FLASH (Foundations of LeArning Systems for alcHemy) group, organized by Tuo Zhao. We are conducting interesting research in machine learning, deep learning, reinforcement learning, natural language processing, etc. We cover a wide range of topics, from inspiring heuristics to serious theories.

What motivated or inspired you to pursue your Ph.D.?

I decided to pursue a Ph.D. back in my undergraduate time, even before I had made my mind to apply to a university in the United States. By then, my major was electrical engineering, and I was doing circuit design and testing things. However, I got a little bored of hardware stuff, and my interest in mathematical driven problems never eclipses.

My fellow students and friends encouraged me to apply for a graduate study abroad — we prepared for the language test and GRE together. That’s why I am very fortunate to do my Ph.D. studies here at Georgia Tech.

What has been challenging, rewarding or unexpected about earning a doctorate degree?

Trying to solve a new problem, or explain a really interesting yet counter-intuitive observation is challenging. This is much more pronounced during the doctorate study, when you are facing seemingly endless possibilities on how interpret a topic. 

The process of unfolding a new topic/problem from the very beginning is rewarding, nonetheless, at the end of day, I still cannot fully understand every bit of it.

I guess the most unexpected thing in my doctorate study is the high difficulty of learning good scientific English writing skills. It is even exaggerated for non-native speakers. I think I am still on the learning curve.

What attracted you to Georgia Tech as the place to earn your Ph.D.?

There are three main reasons:

1) Georgia Tech’s high reputation

2) My advisors

3) Living in Atlanta

My hometown is Hangzhou, and the weather there is pretty similar to Atlanta, except that Hangzhou has more severe high/low temperatures. My advisors are very supportive of my studies and their guidelines not only help in my research problems, but enhance my critical thinking skills and ability to discover problems myself.

What app do you spend the most time using on your phone?

I love YouTube. I watch some random videos, and also learn board games like Go, bridge, and cube skills.

What’s a talent you would most like to have?

I hope I could recover a 3×3 Rubik’s cube in less than six seconds. It would be very cool! Imagine someone spends like 30 seconds to scramble a cube, yet I recover it in 1/5 of the scramble time. Meanwhile, I hope I was able to recover cubes blindly, or by using a single hand.

What is your proudest accomplishment so far?

I have published more than 10 papers.

What’s something someone would be surprised to learn about you?

I can cook Chinese dishes from different regions. I am also experimenting on combining American and Chinese flavors.

If you could time travel to any period of time, where would you go and why?

I would like to restart my undergraduate study with a different major, like math. It took me three years to transition from studying electronic hardware to machine learning. If I were an undergraduate in math, I may have saved a whole lot of time self-studying real analysis and probability. 

Why do you think embodying Georgia Tech’s motto of “progress and service” is important, especially in regards to machine learning (ML) and artificial intelligence (AI)?

Five years ago, I could have never foreseen robots delivering meals by taking the elevator and walking through the corridor by themselves. This is impossible without the innovations in computer vision and robot control brought to life by ML and AI. Scientists and engineers are continuously advancing the frontier of ML and AI, and turning dreams into reality. As we’ll continue to see, the ultimate goal of ML and AI is to better serve people and improve our lives.

Press Contact:

Allie McFadden | Communications Officer |

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.