10.07.20 – The EPFL professor and mathematician won the award for an exceptional research paper that he published.

The IEEE Information Theory Society Paper Award is presented each year by the IEEE Information Theory Society for outstanding research papers in information-related fields such as information processing, transmission, storage and use.

One of the winners of the 2020 award was Professor Emmanuel Abbé, EPFL’s Chair of Mathematical Data Science (SB) and head of the Laboratory of Mathematical Data Science (IC).

“I’m very happy ­– it’s great to have this kind of recognition, especially since there are so many worthy papers out there,” says Abbé. “It’s a major award in this research field, so it is rewarding to receive it.”

Detecting communities

The prize-winning paper is about the stochastic block model, which produces random graphs containing communities and is used in uncontrolled machine learning. The paper proposes an algorithm to determine the fundamental limit for the exact recovery of communities – a problem that had remained unresolved for several decades.

“I started looking at this new approach in 2012, and the award is for a paper I wrote with two former students at Princeton University,” says Abbé. “It was one of the first papers in this line of research.”

See also:

https://actu.epfl.ch/news/the-math-of-detecting-communities/

Due to the coronavirus pandemic, the prize winners were announced in a virtual ceremony that took place on 21 June. The award will be officially presented next year at the IEEE International Symposium on Information Theory (ISIT).