Connaught New Researcher award boosts machine learning research

September 17, 2020

By Tyler Irving

A project from The Edward S. Rogers Sr. Department of Electrical & Computer Engineering (ECE) is one of two U of T Engineering research projects to receive a boost from Connaught New Researcher awards.

The annual awards are only provided to U of T assistant professors within the first five years of a tenured-stream academic appointment to help them establish strong research programs.

Professor Nicolas Papernot (ECE), a recipient of this year’s award, will focus on developing algorithms with his team to make machine learning more trustworthy.

“The widespread adoption of machine learning raises security and societal concerns,” he says. “For instance, learning algorithms can easily be manipulated by adversaries capable of perturbing the data that algorithms analyze.”

By designing machine learning algorithms that can detect and avoid such manipulation, Papernot aims to both increase the algorithms’ ability to learn effectively and responsibly, and to help humans put more trust in machine learning.

“Trustworthiness is instrumental to ensuring a beneficial impact of machine learning,” he says. “We strive to design our experiments in a way that facilitates the translation of our research results into practical techniques and best practices.”

Papernot is among 56 principal investigators across U of T to share more than $1 million in funding through the Connaught New Researcher awards.

“Supporting early-career researchers as they build up their programs is a key priority for U of T Engineering,” says Ramin Farnood, Vice-Dean, Research at U of T Engineering. “I look forward to seeing the exciting technologies that will come out of these innovative projects, and the positive impact they will have on our world.”