5 Questions with Faculty: Siwei Lyu

Associate Professor Siwei Lyu says he enjoys using a team-based learning approach when teaching his students.

ALBANY, N.Y. (Nov. 22, 2017) — Siwei Lyu, an associate Professor of Computer Science at UAlbany’s College of Engineering and Applied Sciences, came to the University years ago, after a three-year post-doctoral stint at New York University.

“I was initially attracted to UAlbany because of the chance to work with prominent faculty,” Lyu said, pointing out in particular Distinguished Teaching Professor Emeritus Sekharipuram S. Ravi. Albany’s central location was also attractive, he said.

What are you working on now?

I am working in the field of machine learning and its applications to challenging real life problems, whether it is about picking out a fake image from the gigantic number of digital images on the Internet and social media, teaching robots to understand the human poses and learn to grasp objects, deciphering brain signals to help patients with injured limbs to recover mobility, or making traffic monitoring cameras “smarter” to recognize crimes or traffic accidents.

What made you decide to pursue your field?

With backgrounds in both computer science and neural science, I am fascinated with the extraordinary ability of the human brain to turn experience and information into knowledge, and I want to make machines also be able to do the same. This research field is the right mixture of mathematics, programming and practical applications that always challenge me to explore more.

What’s your favorite class to teach?

My favorite class is CSI 131: “Introduction to Data Analytics”. This is an entry-level class open to all majors, and I provide them with basics of data analytics and machine learning. I adopt the team-based learning approach for this class, in which students work in small teams to learn together. The team discussion in the class is lively and I am really impressed by the interest and devotion of my students. Every time I teach this class the experience also teaches me something.

What do you see as the challenges confronting those in your field?

Machine learning has experienced tremendous advances in recent years, fueled by huge investments from big tech companies such as Google, Amazon, Facebook and Microsoft. Very impressive performance has been achieved with the current technology with extensive media coverage. However, one fundamental question we still have is, even though computers can “learn,” unlike humans, they cannot explain what they have “learned,” or if they have “learned” the right thing.

This is problematic if results from machine learning are to be used in front of a jury or non-experts that requires interpretation. The whole field should shift and pay attention to the “explainability” of the learned algorithms.

What’s the best thing about working at UAlbany?

One thing I found really appealing about working at UAlbany is we have very broad academic portfolio and faculty with a wide diversity of research interests. This gives me the opportunity to work with colleagues from School of Education or Rockefeller College, and combine my research expertise with applications with real societal impacts.

For more news, subscribe to UAlbany's RSS headline feeds