How a summer research program revived one student’s interest in science

Mary Guiden

Justin Thompson at work in the lab at the University of WashingtonSummer programs are in full swing on the University of Washington campus. That means lots of parents dropping off younger students at “camp” sites in the mornings, groups of students wearing matching t-shirts walking around campus and taking breaks to play Frisbee. Many college-level students will embark on new adventures in a lab.

At our Center, summer means that undergraduates, high school students, teachers and military veterans will have a chance to work in a lab, learn new skills and take what they learn back to the classroom or the next stage in their career.

Last summer, Justin Thompson was one of the first veterans to take part in the Research Experience for Veterans – University Projects (REV-UP) program in Seattle. The program reignited his interest in pursuing graduate work, Thompson said. He’ll start his PhD studies in applied mathematics at the UW this fall.

Thompson took some time recently to talk about his experience. An edited transcript is below.

How did you find out about the REV-UP program?

I got out of the army at the end of 2007 and started school in Olympia, Washington. I went to a community college for two years, transferred to UW and initially, I was double majoring in biochemistry and neurobiology. Then, I switched at the end of my junior year to mathematics. At the end of my senior year, I ran into one of my neurobiology professors, Dr. Michael Kennedy. I explained what I was working on and he said, “So, what are your plans, are you going to grad school?”

I really don’t know what I want to do with the rest of my life. At the time, I was studying pure mathematics, it was very theoretical and very abstract, which was cool at first. But I didn’t see myself doing that for the rest of my life.

I was torn between either going into grad school or possibly teaching. I was also a full-time tutor and I love teaching. So, I had a lot of things on my plate. He said, ‘Well, why don’t you try this out, look into this? The CSNE has a summer research program, they’re looking for veterans, and you could see if you like it.’ So, sure enough, I applied and loved it. It was an amazing experience.

Where did you work at the UW?

I worked in The Daniel Lab in the biology department. The Center paired me with Professor Tom Daniel’s lab because his post-doc, Bingni Brunton, was conducting research that was in line with my interests. Sure enough, that was the perfect pairing for us. They did a great job putting me in with the right group.

What was the coolest thing you got to work on in the lab?

The project that I worked on involved classification of musical genre. I love music and I love math and it was a great blend of the two. It was machine learning but more than anything, I had the chance to learn about the technique of sparse sensing. I’d never heard of it before and as soon as I heard the pitch, my mind started exploding with possibilities. No matter what I do in the future, I will incorporate it into my research.

What is sparse sensing?

Bing and Tom knew that I had an interest in music and they saw that there was a space to develop a machine learning application using music as the data. Sparse sensing is data-driven. It doesn’t matter what your data set is, the algorithms and the methods used can work for any data set. That is one of the key points, and why it’s so useful if you’re a biologist, especially a neurobiologist, looking at biological systems, like a neural network. You have an avalanche of data. There’s so much data, you don’t know what to do with it.

The ideas with sparse sensing is how do we take that data, cut through all of the stuff we don’t need and look at where the information is? So by developing these algorithms for music, it can be generalized and applied to more domains than just music.

What were the results of your research?

Right now, you’re watching my computer screen, and you’re listening to me talk. In the background, you can hear Tom [Daniel] talking. Somehow, you can focus on whatever is of interest. That’s what we call data reduction. With any kind of decision-making process, we want to boil it down to a simple yes or no, a binary decision, A or B.

The primary thing we were looking at is given a sample song, is it in genre A or genre B? First off, we have to turn music into a picture. Basically you do lots of fancy stuff and you end up with these little pictures. This is the average of 50 different blues songs, 50 classical songs and so on. If you can tell the difference visually, we can train the algorithms to also pick up on it.

How much can you whittle down the data?

We’ve thrown away 99.8 percent of the data, and in every case, we got the algorithm to perform better than chance, or better than a coin toss. That was ultimately where this project led.

Looking back, how would you describe the summer research experience?

It changed my mind about everything. I had pretty much abandoned science altogether. I didn’t think that applied mathematics or science was for me at all, but doing this I realized, “Oh wow, there’s a whole realm of academia that I didn’t even know about and that I love.” It changed everything as far as my career path, and my plans for my future.

What happened once you were done?

Tom asked me to stay on in his lab. I applied for graduate school after the summer. Then, I took two months off and went to Thailand with my fiancée. I was accepted into graduate school at the University of Texas at Austin and the University of Washington. I decided in April that I would attend the UW.

Any idea what you’ll do next?

Be happy. (Thompson smiles). Academia is really cool. I like the idea of the freedom. I absolutely love teaching and I could see myself doing that for the rest of my life. But you never know what’s just around the corner. Who knows what interests I’ll find, what options are going to be open until I graduate? I want to keep my options open but I can see myself being a research professor or a professor.

In addition to the REV-UP program, the Center for Sensorimotor Neural Engineering has summer research experiences for undergraduate students, high school students and middle school and high school teachers. Program benefits include an opportunity to be involved in a research project. Participants will gain experience in preparing and presenting research and receive a generous stipend. Application forms for 2015 will be available in November.