In honor of the Department of Statistics' 50th Anniversary, we are highlighting our outstanding faculty and alumni of the past half-century.
Carnegie Mellon Statistics Department alumna Kary Myers (Ph.D. 2006) first came to Carnegie Mellon the summer after her junior year of high school as part of the advanced placement/early admission program. Because of her performance that summer, she earned an invitation to leave high school in Montana early and start her freshman year of college that fall in the Mellon College of Science (MCS). “I declared math as my major and then, over the next three semesters, earned grade point averages of 3 something, 1 something, and 0.00, at which point I was academically suspended and had to go back to Montana,” she said.
When Kary returned to Carnegie Mellon a couple of years later it was as a staff member in the brand new Office of Undergraduate Affairs in MCS. “The position was a perfect fit for me as MCS Dean Susan Henry (now at Cornell) created it to address the large number of MCS students who were failing out (as I did) or otherwise not sticking around until graduation,” Kary said. For eight years she served as assistant to Associate Dean Eric Grotzinger, working with undergraduates and faculty to improve the student experience and reduce attrition. As part of that effort, Kary was asked to analyze data on freshmen throughout the university to see if it could be predicted which students would need help with calculus. Originally, she asked friend Claire Palmgren, who was working for Baruch Fischhoff in Engineering and Public Policy and Social and Decision Sciences, to take on the analysis project. Claire told Kary she didn’t have time to analyze the data herself, but that she could show her how to do it (“which is a pretty amazing statement when you think about it!,” Kary said.) Claire taught Kary how to use SPSS software to start looking at the data, which led Kary to meet with Prof. Joel Greenhouse and learn about logistic regression. “Joel helped me turn the work into a research poster which I presented at the Meeting of the Minds Undergraduate Research Symposium, earning an honorable mention. Then I was hooked!” Kary said.
As a staff member, she could take one or two classes each semester for free. With the support of Deans Henry and Grotzinger, she started chipping away at a bachelor’s degree in statistics with a minor in computer science. In her final few semesters she worked with Profs. Larry Wasserman and Andrew Moore (now dean of the School of Computer Science) on a computational astrostatistics project that turned into her undergraduate honors thesis. “I have supervised one and only one undergraduate thesis student and it was Kary,” Prof. Wasserman said. “She was very independent and seemed more like a graduate student than an undergraduate. “I knew then that she would go on to do great things,” he said.
When she finally completed her degree with university and college honors, Amy Burkert (now the Vice Provost for Education) encouraged Kary to pursue a PhD so that she could ultimately lead her own projects. Kary took her advice, quit her job in MCS, and enrolled in a new joint graduate program between the Statistics Department and what was then the Center for Automated Learning and Discovery (now the Machine Learning Department). Her graduate work was supported by a fellowship provided by AT&T Labs — six years of full funding and a stipend, plus internships at AT&T Shannon Labs in their Artificial Intelligence Department and Machine Learning Department. “I thought I’d end up working there after grad school, but then AT&T’s stock price followed a similar trajectory to my GPA in those first three semesters, leading to big cuts in their R&D budget. “I had to figure out a Plan B. ‘B’ ended up being for ‘Bill Eddy’, and I worked with Bill on my thesis that combined video data of the surface of a brain with simultaneous physiological measurements to identify activation,” she said.
Kary defended her thesis in January 2006, and a few weeks later she was in New Mexico starting her new job as a scientist at Los Alamos National Laboratory, where she is today in the Statistical Sciences Group. Kary said she routinely draws on the interdisciplinary experience she gained in Carnegie Mellon’s Statistics Department. “I’ve studied chemical spectra collected by the Mars Science Laboratory’s Curiosity rover, analyzed electromagnetic measurements to support treaty verification, and developed methods for efficient analysis of the output of huge scientific simulations. “Everything our group does is application-driven, highly interdisciplinary, and often computationally challenging, characteristics that are familiar to any Carnegie Mellon Statistics alum,” she said.
“At Carnegie Mellon I also learned to take an active part in the larger statistics and data science community to bring exciting problems to the attention of a diverse set of experts. “At Los Alamos I created CoDA, the Conference on Data Analysis (cnls.lanl.gov/coda), to showcase data-driven problems of interest to the Department of Energy (DOE) and feature research from the DOE national laboratories as well as from academia and industry. “My goals for CoDA in many ways echo the intentions of Rob Kass’s Case Studies in Bayesian Statistics and Machine Learning workshop series,” she said.