Sindri Magnússon

Sindri Magnússon serves as an Associate Professor in Machine Learning at Stockholm University's Department of Computer and Systems Science. His research focuses on distributed optimization, machine learning, and data-driven decision-making within complex network systems, with applications spanning power grids, IoT infrastructure, and satellite operations.

His educational background includes a B.Sc. in Mathematics from the University of Iceland (2011), an M.Sc. in Applied Mathematics (Optimization and Systems Theory) from KTH Royal Institute of Technology (2013), and a Ph.D. in Electrical Engineering from KTH (2017). He completed postdoctoral research at Harvard University (2018-2019) following a 9-month visiting PhD stint there in 2015-2016.

Magnússon's research integrates theoretical optimization frameworks with practical machine learning implementations, particularly emphasizing reinforcement learning, federated systems, and distributed algorithms for resource-constrained environments. His work bridges theoretical guarantees with real-world applications in critical infrastructure and industrial systems.

Analysis of his recent publications reveals a strong focus on communication-efficient distributed learning, non-stationary reinforcement learning environments, and predictive maintenance systems. His team consistently develops novel algorithms addressing asynchronous updates, model mismatch in federated settings, and multi-objective optimization challenges across IoT and satellite networks.

Major awards include the IEEE ICASSP best student paper award (as supervisor) and a prestigious Swedish Research Council (VR) Starting Grant.

• VR Starting Grant: Resource Constrained Machine Learning in Complex Networks (PI, 4 MSEK, 2021-2024)
• Digital Futures: DEMOCRITUS project on critical societal infrastructures (Co-PI, 4 MSEK, 2021-2024)
• Vinnova: Smart Converters for Climate-neutral Society (PI, 3 MSEK, 2022-2025)

He actively supervises seven PhD candidates including main supervision for Ali Beikmohammadi, Shubham Vaishnav, and Mohsen Amiri, plus industrial PhD projects with Spotify. As Associate Editor for IEEE/ACM Transactions on Networking, he contributes significantly to the networking and machine learning research communities.