'Innovator' spotlight on UW data virtuoso
Abraham Flaxman, a UW professor newly named one of "35 Innovators Under 35," took the abstract mathematics he learned in graduate school and applied it to video games, then turned his talents to global public health.
Seattle Times staff reporter
When high-school math whiz Abraham Flaxman arrived at the Massachusetts Institute of Technology in 1996, he found himself intimidated by how brilliant many of his fellow freshmen seemed. Their math chops, he said, "made me question whether or not this was for me."
But after dabbling in other fields in college, he decided to stick with math. It has certainly paid off.
Last week, MIT's Technology Review magazine named Flaxman, now an assistant professor of global health at the University of Washington, to its annual list of "35 Innovators Under 35."
The distinction — awarded to MIT alumni and non-alumni alike — lands him in the company of technology giants such as Larry Page and Sergey Brin, who started Google; Jonathan Ive, who designed Apple's iPod and iPad; and Mark Zuckerberg, who, of course, is the founder of Facebook.
Flaxman, 34, hasn't created the next Facebook. But his breakthrough also has the potential to change the world, maybe in an even bigger way than Zuckerberg's social network.
For the past four years, Flaxman has fashioned algorithms and models at the UW to try to answer a few big questions: How do you take the available data on the world's diseases and conditions — from malaria to depression — and figure out how many people are contracting, living with and dying from each of them? How do you account for variations over time, by sex and by age? And how do you do that for every country in the world?
Researchers have wrestled with these questions for decades, said Stephen Lim, who works with Flaxman at the UW's Institute for Health Metrics and Evaluation. But Flaxman's deep background in mathematics, uncommon among public-health researchers, Lim said, "brings a unique perspective to the challenges we face."
Flaxman did not set out to work in public health. He grew up in Evanston, Ill., the son of an engineer-turned-lawyer and a psychologist. A summer working in his uncle's medical practice convinced him he was not cut out to be a physician.
After flirting with engineering and economics at MIT, he graduated with a mathematics degree and went off to a Ph.D. program at Carnegie Mellon University in Pittsburgh.
His work in graduate school was, for the most part, highly abstract.
Some of the problems he tackled were based on real-world ones, said Alan Frieze, his adviser at Carnegie Mellon, "but you wouldn't expect them to have real-world applications."
Flaxman, though, has managed to take the theoretical constructs he learned as a grad student and apply them to video games and, later, to global public health.
After Carnegie Mellon, he worked in the Theory Group at Microsoft Research, where he turned his talents to figuring out better ways to match up multiplayer teams on Xbox Live. "You maximize fun if you can't figure out who's going to win," he said with a smile.
As a mathematician with a knack for working with data, Flaxman had a number of career paths open to him after his gig at Microsoft ended in 2008. He could work in finance — not the brightest idea at the time — or on search-engine technology. Or he could work in public health.
"All of these places have interesting puzzles, and it would make a mathematician happy to try to solve some of them," he said. "But the stuff in health, especially in public health, is literally life-and-death issues."
Flaxman won a fellowship at the UW, where he was almost immediately thrown into a massive effort to figure out which populations suffered from various diseases. His part in the project was to build a mathematical machine capable of taking data from a variety of researchers and estimating the prevalence of about 200 diseases in nations around the world.
Some of it, like the HIV data, was very good. But most of it varied wildly. Estimates of how many people have, say, anxiety disorders in a given country can range from 5 to 20 percent. And simply averaging all the numbers together can produce misleading results.
"We've got to know what we don't know," Flaxman said.
Flaxman used a technique called meta-regression to attack the problem. Theo Vos, a professor at the School of Population Health at the University of Queensland in Australia, worked closely with Flaxman and said his efforts have helped estimate the number of people around the world with diseases and conditions from neck pain to emphysema.
Accolades come in, too, from the Bill & Melinda Gates Foundation, where Philip Setel, the deputy director of measurement, learning and evaluation, said that Flaxman's "enormous contribution to the field of health metrics" would help organizations decide how to better allocate their resources.
The improved data Flaxman's models have generated could help the developed world, as well as the developing. It shows, for instance, that Australia has made much more progress in lowering the mortality rate for adults 18 to 65 than the United States has.
Flaxman, suddenly finding himself the center of attention, is more familiar with data and deep thinking.
"I would describe Abie as kind of quietly brilliant," Lim said, using Flaxman's nickname. "He approaches things in a very calm way."
But Flaxman has allowed himself to revel a little bit in being named to the list. "It's super exciting for me," he said. "It's great to have this work recognized."
Theodoric Meyer: 206-464-2985 or firstname.lastname@example.org. Twitter: @theodoricmeyer.