David Tse

"My love of talking and always having an audience, helped cement my decision to become a professor," jokes David Tse.

Passionate about information theory, his research is solving massive problems that incorporate software and mathematics. David was recently announced as the 2017 Claude E. Shannon Award recipient.

What made you decide to be a professor, and what made you want to be at Stanford?

'Professor' is a job in which the audience has no choice but sit and listen. And ever since I was a kid, I've really loved talking. I thought what a perfect way to combine my love of talking and always having a captive audience.

More seriously, my PhD advisor seemed to have a lot of fun being a professor, so I thought it would be a good career.

I've been at Stanford for two and a half years. Before coming here, I taught at Berkeley for 18 years. One reason I came to Stanford is because in recent years, my research is much more interdisciplinary, especially on connecting information theory with biology and genome sequencing. Stanford's campus is a rare find – it is truly interdisciplinary. So not only does Stanford have many top departments, the engineering school is next door to the medical school!

How did you choose your field of research?

When I was in graduate school at MIT, I didn't know anything about information theory until I took a course during my first year. From that point on, I thought the field was amazing! It's abstract, elegant, clean and at the same time has far reaching implications on the designing of communication systems, which are all around us. For example, every time we play a DVD, look up something online, make a call; we are using communication and storage systems all designed on the basis of information theory.

Who has influenced your work and why.

My graduate advisor, Bob Gallager.

As a grad student, when I started working with him as my advisor, I asked him to give me a problem to work on. He refused, saying, "PhD research is more than solving a problem someone gives you. You have to find your own problem. Go read papers, talk to people, come back and we'll talk."

So I read a lot of papers and did a lot of thinking about problems to work on, and every week I would go back to him. And every week he would shoot down all of my ideas. For one and a half years this went on. Of course, I felt like I wasn't producing anything! But in the end, what he taught me was how to uncover a good research problem.

He didn't make me smarter, but what he did teach me makes me a stronger researcher. I think asking the right question is the most important part of research. There are plenty of smart people all around us, but only some ask the right questions that result in making an impact.

Briefly explain a project you are currently working on.

I am working on the world's largest jigsaw puzzles. Using many short DNA subsequences, or reads, my job is to create an algorithm that assembles a complete DNA sequence. This started off as a curiosity-driven project a few years ago: how to abstract a clean mathematical problem that captures the essence of the complex engineering problem of designing assembly software which handles hundreds of Gigabytes of sequencing data. One thing led to another, and now we have designed two assembly software, one for DNA and one for RNA assembly. This journey would not have been possible without a bunch of students and postdocs who are very talented both in theory and in software development.

What advice do you have for new EE students?

A lot of exciting research is connecting the basic disciplinary knowledge we have and applying it to other fields, e.g, biology. Everyone should be on the lookout for interdisciplinary problems. Stanford has many top departments that are interdisciplinary. My advice to a new student is to collaborate with people and take courses that expand your comfort zone. It's good to take advantage of that while at Stanford.

Secondly, remember that there's no huge rush to try and create results. PhD research is 5 years of luxury time to do something interesting. Take the time and find interesting problems to work on instead of working on something immediate.

Don't be scared of messy problems. Try to understand them, abstract their essence and turn then into clean problems. That's what Shannon did throughout his career.

Learning how to do good research – learning what questions to ask – will benefit you whether you go into industry or academia.