Institute of Communications and Connected Systems


ICCS research student receives distinguished award

21 June 2021

Research student Sephora Madjiheurem has recently had her work in collaboration with the company DeepMind celebrated with a distinguished award at the AAAI2021 conference.

AAAI logo in foreground, in background an electronics circuit board with a processor chip embossed with a glowing brain image and AI text

Author: Robert Thompson, Institute Manager

Artificial Intelligence | Reinforcement learning | Industry Collaboration

Sephora Madjiheurem, a research student in the Department of Electronic and Electrical Engineering, and ICCS member, has had her work celebrated with a distinguished award presented at the AAAI2021 conference held earlier this year.

The Association for the Advancement of Artificial Intelligence's conference on Artificial Intelligence (AI) is a top tier conference promoting research in AI and scientific exchange among AI researchers, practitioners, scientists, and engineers in affiliated disciplines.

The work presented was undertaken during an internship with DeppMind where Sephora worked within the Reinforcement Learning Team on a project to design a new algorithm for more efficient information propagation in reinforcement learning settings, where the system learns from trial and error as opposed to learning from labelled data.

The new algorithm developed by Sephora and the DeepMind team significantly improved upon its more traditional predecessor, and the team were able to give theoretical guarantees of the algorithm's performance.

Speaking about the internship with DeepMind and award Sephora said:

"I am very grateful that I had the opportunity to collaborate with researchers at DeepMind, and beyond pleased that our work resulted in an internationally recognised award. I look forward to seeing how the research community will build on this foundational work to further improve efficiency in reinforcement learning."

With all work presented at the seven-day event having to reach the high reviewing standards of the conference, it is an excellent achievement to have work presented at the conference. The distinguished award shows the high calibre of internationally impactful work undertaken by Sephora.

Following her Internship, Sephora is now continuing her research within the Learning and Signal Processing  (LASP) research team at UCL led by Dr Laura Toni.