Outstanding Paper Award for KIOS researchers at the 2025 IEEE Symposium Series on Computational Intelligence

award_ceremony - outstanding paper

The KIOS research team, consisting of PhD Student Charalampos Shimillas, Senior Research Associate Dr. Kleanthis Malialis, PhD, and Professor Marios Polycarpou, together with Professor Konstantinos Fokianos from the Department of Mathematics and Statistics at the University of Cyprus, has received the Outstanding Paper Award for their research work on “Transformer-based Multivariate Time Series Anomaly Localization”.

The awarded paper addresses the challenging task of localizing anomalous time series in large-scale cyber-physical systems like critical infrastructure systems, a crucial step in ensuring system reliability and safety. To address this, the research introduces a new method that uses transformer models, powerful state-of-the-art AI methods, to analyze complex patterns in multivariate time series data. By analyzing how these models focus on important details over time, the proposed approach improves the accuracy of finding unusual system behaviors. This work enhances the ability to detect and pinpoint faults in critical infrastructure systems, helping to maintain safety and reliability.

This paper was presented at the 2025 IEEE Symposium Series on Computational Intelligence (IEEE SSCI 2025) in Trondheim, Norway. IEEE SSCI is widely recognized for cultivating the interchange of state-of-the-art theories and sophisticated algorithms within the broad realm of Computational Intelligence Applications.

This research work was supported by the European Union’s Horizon Europe research and innovation programme under grant agreement No 101073307 (MSCA-DN LEMUR), the European Research Council (ERC) under grant agreement No 951424 (ERC-SyG Water-Futures), the European Union’s Horizon 2020 research and innovation programme under grant agreement No 739551 (Teaming KIOS CoE), and the Republic of Cyprus through the Deputy Ministry of Research, Innovation and Digital Policy.

Link to the paper:  https://arxiv.org/abs/2501.08628