Sindri Magnússon Senior Lecturer, Associate Professor

I received the B.Sc. degree in Mathematics from the University of Iceland, Reykjavík, Iceland, in 2011, the M.Sc. degree in Applied Mathematics (Optimization and Systems Theory) from KTH Royal Institute of Technology, Stockholm, Sweden, in 2013, and the Ph.D. in Electrical Engineering from the same institution in 2017. I was a postdoctoral researcher at Harvard University from 2018 to 2019, and a visiting Ph.D. student at Harvard for nine months in 2015–2016.

My research is broadly in artificial intelligence and machine learning, with an emphasis on data-driven decisions and operations in complex systems such as cyber-physical and socio-technical networks. This includes the development of methods and theory for learning, control, and optimization that enable reliable and efficient operation of large-scale interconnected systems. Within this broader context, I study how multiple agents can learn and adapt in cooperative settings, developing algorithms for multi-agent learning, distributed decision making, and reinforcement learning. Building on this foundation, my research has more recently expanded to fairness and bias in AI decision making, exploring how algorithmic choices can inadvertently disadvantage certain groups and how principled methods can be designed to mitigate these effects. By combining learning theory with system-level applications, I aim to build AI that is scalable, efficient, and socially responsible.

I have extensive experience as a project leader, including as Principal Investigator of major research projects such as the Swedish Research Council (VR) Starting Grant Resource Constrained Machine Learning in Complex Networks (4 MSEK), the VR Project Grant Federated Reinforcement Learning: Algorithms and Theoretical Foundations (4 MSEK), and the Vinnova project Smart Converters for Climate-neutral Society: Artificial Intelligence-based Control and Coordination (7 MSEK). As a supervisor, I received the Best Student Paper Award at IEEE ICASSP 2019. I currently serve as an Associate Editor of IEEE/ACM Transactions on Networking and have been a Technical Program Committee member for multiple leading conferences in control and machine learning, including IEEE INFOCOM.

I have authored over 70 peer-reviewed publications, including work in top AI and machine learning venues such as NeurIPS, ICML, AAAI, and Transactions on Machine Learning Research, as well as widely recognized journals and conferences in signals, systems, and communications, and in AI-driven operations in distributed systems and cyber-physical systems, including energy systems and the Internet of Things (IoT).

Mentoring is central to my academic work, both as a responsibility and as a source of inspiration. I have supervised more than 30 master’s theses in data science, artificial intelligence, and related areas. At the doctoral level, I have been fortunate to mentor several very talented PhD students, serving as main supervisor for Ali Beikmohammadi, Shubham Vaishnav, Mohsen Amiri, Guilherme Dinis Junior, and Alireza Heshmati, and as co-supervisor for Lida Huang (graduate in 2025), Zahra Kharazian, Sayeh Sobhani, Alfreds Lapkovskis. Their projects span topics such as multi-agent learning, federated learning, predictive maintenance, reinforcement learning, and fairness and bias in AI decision making, with applications in socio-technological and cyber-physical systems. Together, these contributions advance both the theoretical foundations of AI and its practical applications in complex, real-world domains.