Machine-learning expert Jegelka wins major German award

October 17, 2015

Adam Conner-Simons | MIT CSAIL

Stefanie Jegelka wins “Deutscher Mustererkennungspreis” prize. 

Image: Stephanie Jegelka

This week CSAIL principal investigator Stefanie Jegelka received a prestigious German prize for her research efforts in machine learning.

Considered the highest recognition awarded by the German Pattern Recognition Society, the “Deutscher Mustererkennungspreis” prize is given to an outstanding young researcher in machine learning or computer vision under the age of 35. Jegelka is the first woman to receive the award.

Jegelka, who joined MIT earlier this year as an assistant professor, was honored for her work on submodular optimization for machine learning and computer vision applications.

Jegelka says that recognizing that a problem has submodular structure can be extremely beneficial.

“It often means that good algorithms may be available for solving the problem, and beautiful mathematical structure can be used,” she says.

Her work has demonstrated practical, fast and widely applicable algorithms for solving problems with submodular properties. For example, Jegelka was able to design a method for finding complexly-shaped objects in images by incorporating prior knowledge about regularities that help identify object contours.

“Such algorithms are not only easy to work with, but also come with theoretical guarantees about the solutions they find,” she says. “Our algorithms have also been used for tasks as diverse as computing confidences about inferences in probabilistic models, optimizing topologies of social networks, or effective exploration.”