Community has always been important to Sarah Cannon. The School of Computer Science (SCS) Ph.D. student chose Georgia Tech for its encouraging and collegial environment. Now in her final year, she’s involved in theory group research, grad student social groups, and a push to get more women in CS.
“The types of things I'm interested in are right on the border of math and computer science, and I could do both at Tech in a friendly and supportive environment,” she says.
Although Cannon is now a distinguished student in the Algorithms, Combinatorics, and Optimization (ACO) doctoral program, she first discovered theoretical computer science in her sophomore year at Tufts University. The math major was getting a minor in CS when she took an algorithms class and excelled. After the semester, her professor asked her to join a computational geometry research project focusing on graph orientations. The experience pushed her to pursue a career in computer science, leaving Tufts with an acceptance to ACO.
But before the New Hampshire native moved South, she spent a year earning her master’s in Mathematics and the Foundations of Computer Science at the University of Oxford. The experience led her to refocus specifically on algorithms when she arrived at Tech in the fall of 2013.
Cannon studies the mathematical foundations of Markov chains for discrete geometry problems, specifically with applications in statistical physics and distributed computing. Markov chains are algorithms used for generating random samples.
“You can accomplish a lot more if you’re okay with things not happening 100 percent of the time but 99.9 percent of time,” Cannon says. “Randomized algorithms enable things to be a lot simpler and faster.”
With a Markov chain, small local changes can dramatically affect the global behavior of an algorithm. Once this effect is understood, it can be harnessed to accomplish new research goals.
Cannon has had the opportunity to see the impact of her research firsthand through her work with the College of Physics’ Complex Rheology And Biometrics (CRAB) Lab. One focus of this collaboration is distributed algorithms for robot swarms, where the goal is for the robot swarm as a whole to exhibit interesting behavior. This is accomplished by carefully defining local algorithms. Using Markov chains to develop these local algorithms allows the researchers to prove what would happen and what behavior the swarm would exhibit.
“We wanted to take our theoretical knowledge, the intuition we have from working with these problems, and apply it to explain or motivate some of these phenomena the robotics researchers were seeing,” Cannon says.
Building a Community
Cannon credits her research opportunities to the tight-knit theory lab and her advisor, Professor Dana Randall. “She always has so many interesting problems she’s bringing to my attention, and she always gives me the freedom to choose what I want to work on.”
Yet Randall also encourages students to have a life outside of their studies. After nearly five years, Cannon has found a balance captaining the Atlanta women’s field hockey team, running, and reading Agatha Christie mysteries before bed.
That support, both in studies and extracurricular activities, is something Cannon found unique to Tech, and she’s tried to bring some of that spirit to the student community. Cannon organizes a weekly graduate student happy hour, and she has also made it one of her goals to help increase diversity in CS.
When Cannon first arrived at Tech, she was the only woman in her 10-person ACO cohort. To connect to more women in the college, she has been on the organizing committee of Georgia Tech Graduate Women in the College of Computing since 2015. The group hosts biweekly coffee breaks to get to know each other, a semesterly faculty lunch with female professors, craft nights, happy hours, and more. It’s even become an informal mentorship group as students can discuss problems and get advice from each other.
“It’s about connecting to the female grad students here, getting to know each other, and having a support network,” Cannon says.
Preparing for Life After Tech
In her time at Tech, Cannon has had quite an impact. She was a 2015 recipient of the Simons Award for Graduate Students in Theoretical Computer Science, awarded to 10 or fewer students at U.S. and Canadian universities each year. She is also an NSF Graduate Research Fellow and was previously a Clare Boothe Luce Outstanding Graduate Fellow.
Her success at Tech has prepared her well for her future career in academia. She partially chose Tech for the opportunity to be able to both teach and do research. Last summer, she was an instructor of record for an undergraduate algorithms course. Her depth of algorithms knowledge enabled her to pare down what would be most relevant to the students. The most rewarding moments were when everyone connected with the material.
“To get students in a required course excited about something can be difficult,” Cannon says. “But there were a few times I presented something that got the students to say, ‘Wow, that’s cool!’”
As Cannon applies for jobs that will enable her to do this kind of teaching in addition to her research, she is still active in the Tech community. This January, she will be attending the Symposium for Discrete Algorithms (SODA) in New Orleans and presenting at the Joint Mathematics Meeting (JMM) in San Diego. This is all on top of working on her thesis combining all of her theoretical research.
Cannon calls herself a theorist fundamentally, but just like her research with robots connects theory to the real world, she’s helped to connect a community at Tech and show just how much a woman in computer science can do.