Balcan Named 2011 Microsoft Research Faculty Fellow


July 17, 2011

Assistant Professor Nina Balcan of the School of Computer Science has been awarded a 2011 Microsoft Research Faculty Fellowship to pursue her work addressing foundational questions in machine learning, algorithmic game theory and optimization.

The Microsoft Faculty Fellowship program is intended to identify early in their careers future research and thought leaders—those who “are exploring breakthrough, high-impact research that has the potential to help solve some of today’s most challenging societal problems,” according to Microsoft. Balcan is one of just eight fellows worldwide for 2011; she and her fellow recipients were formally announced at the 12th annual Microsoft Faculty Research Summit, being held July 18-20 in Redmond, Wash.

Each chosen fellow receives a cash gift as well as software, invitations to academic and professional conferences, and the opportunity to engage with leading researchers from Microsoft Research. Fellows have the freedom to plan their research agenda, hire grad students, build labs and purchase equipment.

“One direction I am very excited about is developing algorithms and basic principles for interactive machine learning methods,” Balcan said of the research her fellowship will support. “Most of machine learning methods so far have been based on passive learning: a human expert provides labeled examples (e.g., for spam detection, this would be email messages and whether they are spam or not) and then the algorithm aims to find the best rule it can from this data, but this can require the human expert to label a lot of data if the rule being learned is at all complicated.

“In my work we show how an interactive approach where the algorithm and human expert interact on which examples to be labeled can dramatically cut down on the work the human expert has to do,” she continued. “This can potentially allow machine learning to be used in many more areas. This Microsoft fellowship will allow me to more deeply investigate these core problems, by providing funds for student support, collaborative visitors and postdocs.”