Graduate Student Spotlights
David Rushing Dewhurst - Complex Systems and Data Science
Area of study: Ph.D. Complex Systems and Data Science
Faculty advisor: Peter Dodds, Chris Danforth, Brian Tivnan
From: Massachusetts
Why did you choose UVM? What do you like best about being at UVM?
I actually was admitted to UVM as an undergraduate to study music performance—a little different from my current field of study, though not so different as one might suspect. I received my undergraduate degree (not in music performance!) and master’s of science at UVM and greatly enjoyed working with my advisors, so I decided to stay here for my Ph.D.
Why did you choose this area of graduate study? What career will you pursue when you complete your degree?
Data science draws together insights from many fields—physics, computer science, mathematics—and presents us with a nearly-universal set of tools with which we may study disparate phenomena. Upon graduation I will continue my career as a research scientist.
Tell us a little about the research or project you are working on.
I am mostly interested in stochastic processes and time series, with particular application to financial economics. Currently, I'm uncovering connections between a type of growth process and mechanistic generation of extreme values in asset return time series.
Can you share a time when a faculty advisor or mentor connected you with either new insights or a valuable opportunity (conference, publication etc.)?
I've been fortunate enough to have many of these experiences. To name a few: my advisor (and good friend), Peter Dodds, connected me with Brian Tivnan, who now serves as my co-advisor; Chris Danforth introduced me to the Mathematical Sciences Research Institute and helped me apply to one of their summer schools, to which I eventually went; and Brian Tivnan has connected me with several stellar career opportunities.
Have you had an internship? If yes, where was it? What were your responsibilities? What were the highlights?
I've had a few internships between undergraduate and graduate school put together. When I was an undergraduate at UVM, two of my majors were in economics and political science, so my internships were largely focused in that realm. I worked at the Massachusetts state house one summer, and subsequently worked in Washington for the Cato Institute and the Tax Foundations, where I studied and wrote on financial regulation and tax policy respectively. As a graduate student, I've been fortunate to work every summer with one of my advisors, Brian Tivnan, and two of my close friends and fellow graduate students, Colin Van Oort and John Ring IV, at the MITRE Corporation, a not-for-profit company that operates federally-funded research and development centers.
Papers
Dodds, Peter Sheridan, David Rushing Dewhurst, Fletcher F. Hazlehurst, Colin M. Van Oort, Lewis Mitchell, Andrew J. Reagan, Jake Ryland Williams, and Christopher M. Danforth. “Simon's fundamental rich-get-richer model entails a dominant first-mover advantage.” Physical Review E 95, no. 5 (2017): 052301.
Dewhurst, David Rushing, Christopher M. Danforth, and Peter Sheridan Dodds. “Continuum rich-get-richer processes: Mean field analysis with an application to firm size.” Physical Review E 97, no. 6 (2018): 062317.
Dewhurst, David Rushing. “Some results on a class of functional optimization problems.” arXiv preprint arXiv:1804.00087 (2018). (M.S. thesis, not published but elements of it form a paper in preparation)
Kristin McClure - Complex Systems and Data Science
Area of study: Complex Systems and Data Science, Master’s Degree
Faculty advisors: Dr. Peter Dodds, Dr. Chris Danforth
From: South Burlington, Vermont.
Why did you choose UVM? What do you like best about being at UVM?
It really came down to two things for me: quality and culture. UVM’s Complex System and Data Science program is leading edge. The leadership team designed a program that blends the intersection of computer science, statistical insight, and communication in an outstanding and innovative manner. Based on the design of the program, I know I will acquire and master critical skills that will help solve small and large problems of our future. In addition to the quality of the program, the culture of the program is one of inclusiveness, high performance, and collaboration.
Why did you choose this area of graduate study? What career will you pursue when you complete your degree?
I started my career at IBM—applying the supply and demand principles to business applications. And as the supply chain matured, it evolved into a data rich environment. Mining the data, analyzing the data, observing patterns and trends, and being able to clearly and concisely tell a story from that data had power, and I knew that was what I wanted to do with my career. I wanted to tell stories from data that haven’t been told before. I wanted those stories to be precise and comprehensive and for those stories to improve lives and communities. With my experience and master’s degree in Data Science from UVM, my goal is to lead a data scientist team focused on improving our world for individuals and communities, while also improving the effectiveness of the organization that provides these services.
Tell us a little about the research or project you are working on.
My machine learning professor recently connected me with an exciting opportunity to partner with a senior member of the environmental engineering team. We are working together to develop a machine learning algorithm that can predict water quality. The applications would provide real-time feedback on water quality with the ability to guide water management practices, policies, and applications.
If you had a time machine and could go back to before you started your graduate program, what advice would you give yourself?
I came from industry and I think one of the most important things you can do for your career – whether it is in academia or industry, is to keep learning. And as such, I am really glad I took the leap to invest in my career with my studies at UVM for my masters. But it hasn’t always been an easy road! And if I had a time machine, I would tell myself the following:
1. Plan it out. I’m a planner by nature and prior to my program starting I met with my advisors and sketched out a path – based on my career goals and areas of interest. This has proven to be very valuable.
2. Math is not like riding a bike. If you are lucky, the high level math will come back to you. But if you are like me, after doing years of basic math at work, it won’t. Brush up on it.
3. DO IT, if you love it, do it! I wish I’d started my program earlier!
Awards
I was awarded a patent for developing an inventory management system that optimizes inventory consumption while working for IBM.
Sage Hahn - Complex Systems and Data Science
Area of study: MS Complex Systems and Data Science
Faculty advisor: Safwan Wshah
From: Concord, Massachusetts
Why did you choose UVM? What do you like best about being at UVM?
A big reason I chose to pursue a master’s at UVM after having completed my undergraduate here was because of the incredible connections I have made with faculty and other students. The various people I have met and continue to work with are by far my favorite part about UVM, in particular Alan Ling!
Why did you choose this area of graduate study? What career will you pursue when you complete your degree?
My undergraduate degree was in computer science, but I have always been a strong proponent for interdisciplinary study. The complex systems program offers a great mix of theory and practice, as well as exposure to a huge expanse of different ideas. While the prospect of choosing a career at this point in my life does not produce quite as much dread as it used to, I am very glad to still have a year or two to decide. That being said, I have recently grown very interested in the intersection between medicine and computer science.
Tell us a little about the research or project you are working on.
My main research project at the moment is in collaboration with the Vasculur Surgery Program at the UVM Medical Center, working toward the automated detection of endoleaks. Endoleaks, of course, are perigraft blood flow outside the endograft post endovascular abdominal aortic aneurysm repair—which is just a lofty way of saying detecting little circles of contrast within bigger circles of contrast. I am also interested in the application of deep learning AI techniques on other medical imaging tasks, such as detecting spinal fractures, and am looking forward to engaging with new ones.
Can you share a time when a faculty advisor or mentor connected you with either new insights or a valuable opportunity (conference, publication etc.)?
A big shout out to Safwan Wshah here, who was responsible for both encouraging me to apply to graduate school and connecting me with everyone over at UVMMC who I am currently working with. I’m proud to be a member of his new research group Vermont AI Lab (VAIL), and greatly look forward to continuing to work with them.
Papers and Awards
I have not been published yet, but am submitting two papers to the IEEE International Symposium on Biomedical Imaging:
• “Automatic Deep Learning Based C2 Spinal Fracture Diagnosis”
• “Deep Learning for Detection of Endoleak after Endovascular Abdominal Aortic Aneurysm Repair”
I was also recognized by UVM as an Outstanding Computer Science Senior.
Shawn Beaulieu - Complex Systems and Data Science
Area of study: Ph.D., Complex Systems
Faculty advisor: Nick Cheney
From: Colchester, Vermont
Why did you choose UVM? What do you like best about being at UVM?
The people I have the privilege of working with are what drew me here. Everyone here is incredibly patient and nurturing with their students, and very shrewd in their research. They’re powerfully driven, and derive real pleasure seeing one another succeed. It’s inspiring to see and take part it.
Why did you choose this area of graduate study? What career will you pursue when you complete your degree?
The program director, Dr. Peter Sheridan Dodds, likes to say that this field can be described as the “study of interesting things.” Complex systems encompass everything that emerges from the interactions of simple (if otherworldly) constituents that are, by themselves, unremarkable. This includes brains, language, robots, energy systems, markets, climate, and more—as well as their intersection. Or, more succinctly, it concerns everything we actually care about.
Tell us a little about the research or project you are working on.
I’m working to develop algorithms that move so-called artificial intelligence systems beyond the brittle and myopic simulacra of intelligence they currently are (which are ripe for abuse by governments and corporations) and toward the dynamic, self-actualizing creatures we see in nature.
If you had a time machine and could go back to before you started your graduate program, what advice would you give yourself?
Don’t be so hard on yourself. I’m reminded of the somewhat tawdry lyric by the band IDLES on their song “Television:” “If someone talked to you the way you do to you, I’d put their teeth through.” Also, get more sleep and make time to go outside.
Papers
My first paper, and the core of my master’s thesis, “Combating catastrophic forgetting with developmental compression,” was published earlier this year at the Genetic and Evolutionary Computation Conference (GECCO). It explored the ways in which Baldwinian evolution, an emergent process of Darwinian evolution under certain conditions, can discover robot controllers (i.e. neural networks) that are better able to avoid what’s called ‘catastrophic interference’—that is, competition between tasks such that improvements on one degrade performance in another—than controllers obtained by other means. This was published with my advisor at the time, Dr. Josh Bongard, and fellow Ph.D. student Sam Kriegman.
Sandhya Gopchandani - Computer Science
Area of study: Computer science (CS), master’s of science (MS)
Faculty advisor: Chris Danforth
From: Pakistan
Why did you choose UVM? What do you like best about being at UVM?
To be honest, I did not know much about Vermont before coming here but when I was doing my research for MS programs in the U.S., I really liked the research being done in the computer science and complex system departments. Moreover, I got full funding in the form of a GTAship. So, that strengthened my decision to come here.
Why did you choose this area of graduate study? What career will you pursue when you complete your degree?
I always see computer science as a medium to solve problems around me. It gives me the opportunity to be creative and helps me come up with actual solutions to real world problems. Also, CS has become a fundamental skill and can be applied to any other field.
Tell us a little about the research or project you are working on.
My current research revolves around social media and mental illness. We want to analyze social media and find clues and patterns about individual’s mental health from their behavior on social media. The research involves data science pipelines, neuro-linguistic programming (NLP) and machine learning (ML) techniques.
If you had a time machine and could go back to before you started your graduate program, what advice would you give yourself?
I would advise myself to reach out to people and make connections. Do not wait for others to initiate conversations. Something good happens when you go out of your comfort zone.
Can you share a time when a faculty advisor or mentor connected you with either new insights or a valuable opportunity (conference, publication etc.)?
Yes, there have been many times when I got to know about conferences and sessions happening both in the university and outside the university from my professors and peers. One of the conferences is Grace Hopper, which I learned about from my advisor last year. Moreover, I was asked to join lunch sessions with potential faculty to assess and provide my feedback. Such opportunities helped me share my opinions and learn from other experiences.
Have you had an internship? If yes, where was it? What were your responsibilities? What were the highlights?
Yes, I was a research and development summer intern at iManage at Chicago. I was working in a team responsible for bringing AI into its existing solutions. So, I had an opportunity to work with various innovative ML and NLP projects. My work involved getting data, preprocessing it and building ML models for tasks such as email classification and predicting high-risk employers. One interesting project that I worked on was to come up with minimum viable product to test the possibility of integrating continuous speech interface in its main product. So, I developed Alexa skills using Amazon open-source application programming interfaces (APIs) that communicated with its existing system. The company demoed this to multiple clients and the response was very positive. All three projects are currently in production or will be soon. The highlight of the internship was working on real projects with people who were very helpful and cooperative.
Awards
I was one of the 220 out of 6,000 people selected to participate in UGRAD academic cultural semester exchange program at University of Wisconsin Eau Claire in Fall 2014. The program was fully funded by the U.S. Federal Government to promote cultural exchange between the U.S. and Pakistan. Moreover, our final year undergraduate project won third place. The project included a mobile application and micro controller Arduino that would enable people with disabilities to operate their appliances using only their facial expressions.