Narayan, Martins participate in BIRS decentralized systems workshop

news story image

Professor Prakash Narayan (ECE/ISR) and Associate Professor Nuno Martins (ECE/ISR) were invited participants at the Workshop on Optimal Cooperation, Communication, and Learning in Decentralized Systems, Oct. 12–17.

The workshop was held at the Banff International Research Station for Mathematical Innovation and Discovery (BIRS) in Alberta, Canada. It addressed decentralized decision-making, a unifying theme of a very diverse set of problems and applications of current interest both to researchers and to practitioners. Some of the problems include collective decision-making and problem-solving via crowdsourcing and social networks; design and implementation of economic mechanisms and control policies for smart grids, health care, and other large-scale socioeconomic infrastructures; parallel and distributed computing for large-scale machine learning problems; and cybersecurity.

Researchers are increasingly distilling a common set of core features shared by these heterogeneous problem domains, and identifying a common list of core questions that need urgent attention. This workshop brought together researchers from communities such as stochastic control, economics, machine learning, and information theory, that do not normally interact. It highlighted the common threads and methodological approaches underlying much of the research into large-scale decentralized systems.

Each of the 41 participants gave a presentation on the theme as part of the event. Martins presented “Distributed estimation over shared networks: optimal event-based policies,” while Narayan’s lecture was on Interactive multiterminal communication.”

BIRS is a joint Canada-US-Mexico initiative that was started in 2003. Every year, the station holds more than 70 different programs for some 2,000 researchers from 400 institutions in more than 60 countries. The workshop that Narayan and Martins attended is one in BIRS’ annual series of 49 five-day workshops, each hosting up to 42 researchers in disciplines in which mathematics, computer science and statistics are used in novel ways. The format allows scientists to exchange the latest advances in their fields of study and provides an environment that fosters new collaborations and ideas.

These workshops address the imperatives of collaborative and cross-disciplinary research with a focus on the mathematical sciences and their vast array of applications in the sciences and in industry. They are designed to facilitate intense and prolonged interactions between scientists in a secluded environment, allowing for uninterrupted research activities in a variety of formats. BIRS embraces all aspects of the mathematical, computational and statistical sciences, from the most fundamental challenges of pure and applied mathematics, theoretical and applied computer science, statistics, and mathematical physics, to financial and industrial mathematics, as well as the mathematics of information technology, and the life sciences.

Published October 27, 2014