PhD Student Pawan Poojary Wins Best Student Paper Award at 2023 WiOpt Conference

Northwestern Engineering’s Pawan Poojary earned the Best Student Paper Award at the 21st International Symposium on Modeling and Optimization in Mobile, Ad hoc, and Wireless Networks (WiOpt) last month in Singapore.

Poojary is a PhD student in electrical engineering and a member of the Communications and Networking (Commnet) Laboratory in the Department of Electrical and Computer Engineering. His research focuses on analyzing multi-agent learning and decision-making problems arising from social interactions using applied probability, learning, and game theoretic tools.

The annual WiOpt conference showcases state-of-the-art, theoretical, experimental, and empirical research related to the modeling, performance evaluation and optimization of networks.

“I feel greatly honored to receive this prestigious award. I am glad that the award jury found my work to be an impactful contribution to the literature,” Poojary said. “I hope that this recognition will shine the spotlight on my research in social learning and will create opportunities for future collaborations that could help tackle such real-world research problems.”

The winning paper, titled “Welfare Effects of Ex-ante Bias and Tie-breaking Rules on Observational Learning with Fake Agents,” is co-authored by Poojary’s adviser, Randall Berry, John A. Dever Chair of Electrical and Computer Engineering at the McCormick School of Engineering.

In this work, Poojary and Berry consider a sequential observational learning setting, such as an online market, where buyers seek to learn the quality of an item up for sale from observations of other buyers’ actions.

“One would expect that if there exist some ‘fake’ buyers that degrade the quality of observations, then this would result in reduced welfare for the ‘honest’ buyers,” Poojary said. “Our main result shows that, under certain conditions that arise from a bias in the buyers’ preferences and their choice of a tie-breaking rule, a counter-intuitive phenomenon occurs — the presence of fake buyers instead leads to a gain in the welfare of honest buyers. Thus, in this field of learning, exploring the relation between the corruption of observations and buyers' welfare merits further attention.”

As a next step in the research, Poojary and Berry aim to determine sharp characterizations of instances when degraded observations lead to lower welfare. They also plan on examining additional forms of noisy observations.

Poojary earned a master’s degree in electrical engineering from the Indian Institute of Technology Madras co-advised by Krishna Jagannathan and Sharayu Moharir. He received a bachelor’s degree in electronics engineering from the Vidyalankar Institute of Technology at Mumbai University in Mumbai, India. Following the completion of the PhD program, he plans to seek a post-doctoral position in industry or academia focused on research in emergent techniques in data-driven, multi-agent learning.