As a teenager in the Czech Republic, Lenka Zdeborová glimpsed her future in an Isaac Asimov novel. A character in Asimov’s “Foundation” series invents a mathematical method for predicting the path of an entire civilization by averaging out the haphazard behavior of billions of individuals. The concept gave her a “mesmerizing feeling,” Zdeborová recalls — one that returned when she later encountered a method that she could actually apply to make sense of huge numbers of unpredictable elements.
“I realized, ‘Oh my God, Asimov was just describing statistical physics,’” she said, referring to a discipline that describes the big-picture properties of matter by using the rules that apply to individual molecules. As a physics master’s student at Charles University in Prague, she reveled in its predictive power. Then, while she was pursuing her doctorate, Zdeborová’s adviser showed her a paper that applied the techniques of statistical physics to theoretical computer science — the mathematical study of computation and how algorithms behave. The familiar feeling returned with a vengeance.
“I was completely mesmerized by that paper,” Zdeborová said. “I had always had this impression that to do computer science, you had to be a hacker and know everything about Linux. I realized that theoretical computer science was as fascinating as theoretical physics, and I said, ‘OK — this is what I want to do.’”
Zdeborová now leads the Statistical Physics of Computation Laboratory at the Swiss Federal Institute of Technology Lausanne. Her work currently focuses on how the physics of phase transitions in matter — such as water freezing into ice — can help model the behavior of algorithms, especially those used in machine learning.
