With the successful defense of her doctoral dissertation, titled “A Federated Learning Framework for Explainable Proactive Task Offloading in IoT Systems” doctoral candidate Klea Elmazi has officially completed her Doctoral Study Programme.
This dissertation focuses on managing computational and communication resources in Internet of Things (IoT) systems, with a specific emphasis on proactive task offloading. The research proposes an integrated framework that combines predictive digital twins, federated multi-agent reinforcement learning, explainable artificial intelligence methods, and reasoning assisted by large language models (LLMs). The primary objective of the research is to enable IoT devices to make highly adaptive task-scheduling decisions that optimize energy consumption, computational load, and communication latency, while simultaneously ensuring decision-making transparency and data privacy. The results validate the effectiveness of integrating predictive modeling, distributed learning, and explainable decision-making into a unified resource management framework for distributed IoT systems.
The doctoral dissertation was supervised by Prof. Jonatan Lerga (Faculty of Engineering, University of Rijeka).
The Doctoral Dissertation Defense Committee consisted of: Prof. Ivo Ipšić (Chair), Prof. Sandi Ljubić (Member) – both from the Faculty of Engineering, University of Rijeka – and Prof. Krešimir Pripužić (Member) from the Faculty of Electrical Engineering and Computing, University of Zagreb.
Congratulations to our colleague!
