ACG 150

The members of the AI Research Team perform and pursue research in machine learning; big data mining; machine learning on sequential data; explainable and trustworthy AI; recommender systems; rule-based systems; natural language processing (NLP) and large language models (LLMs); deep learning in vision; optimization; complex graph analysis; computer game AI, etc. Particular focus application areas include: quality management in Industry 4.0; diagnosis, prognosis and management of hematological diseases; drug-drug interaction prediction; drug re-purposing; social media analytics; pattern mining in biomedical text; machine vision for medical informatics; fault detection and selection in power transmission lines; security incident detection and response in smart grids and others.

The AI Research Team is actively pursuing various research grants from programs such as the European Commission’s Horizon program, the European Space Agency, and others, in collaboration with several external partners.

AI Research Team Members

Currently, the AI lab has 5 members: Dr. Ioannis Christou, Dr. Ioannis Vetsikas, Dr. Elena Chatzimichali, Dr. Dimitrios Vogiatzis and Dr. Nikolaos Bakas.

Mission

To conduct high-caliber and high-impact research in select AI areas; to attract funding through proposal writing; to attract and retain students to do research in AI; to collaborate with other research teams on inter-disciplinary topics and applications related to AI; to foster collaborations with academia and industry.

Vision

To perform world-class research in AI and its applications.

Resources

The AI team has at its disposal the following computing resources:

a. Server with 9.82 TB disk, 2 Xeon E5 v4@ 2.4, 40 Cores, 128 Gbytes, 8 Network Cards and an NVIDIA 12GB Accelerator.

b. 10,000 CPU hours available on the Meluxina European Super-Computer between September 2023 – March 2024 for research and teaching purposes.

Projects

FIREMAN (2019-2023)

Using Machine Learning for Industry 4.0 use-cases

FIREMAN, which was recently successfully completed, was funded by the EU CHIST-ERA program. The AI team, led by Professor Ioannis Christou, developed ML-based algorithms for fault prevention and collaborated with researchers from RTIN’s SWIFT Lab, which developed techniques for wireless data acquisition.

ACG Scorer (2022-2023)

Automatic creation of student study plans at ACG

ACG SCORER is a user-friendly tool that allows students and program advisors to create personalized study plans for the students that optimize a number of complex criteria, including time-to-completion, expected grade point average, etc. The tool uses advanced Quantitative Association Rule Mining as well as mathematical programming techniques to find the optimal solution to a highly constrained multi-criteria stochastic optimization problem.

Select Recent Publications

[1] Daniel Gutierrez-Rojas, Ioannis T. Christou, Daniel Dantas, Arun Narayanan, Pedro H. J. Nardelli, Yongheng Yang: Performance evaluation of machine learning for fault selection in power transmission lines. Knowl. Inf. Syst. 64(3): 859-883 (2022).

[2] Ioannis T. Christou, Evgenia Vagianou, George Vardoulias: Planning Courses for Student Success at the American College of Greece. CoRR abs/2207.02659 (2022)

[3] Alexander Beattie, Pavol Mulinka, Subham Sahoo, Ioannis T. Christou, Charalampos Kalalas, Daniel Gutierrez-Rojas, Pedro H. J. Nardelli: A Robust and Explainable Data-Driven Anomaly Detection Approach For Power Electronics. CoRR abs/2209.11427 (2022).

[4] Daniel Gutierrez-Rojas, Mehar Ullah, Ioannis T. Christou, Gustavo Matheus de Almeida, Pedro H. J. Nardelli, Dick Carrillo, Jean Michel de Souza Sant’Ana, Hirley Alves, Merim Dzaferagic, Alessandro Chiumento, Charalampos Kalalas: Three-layer Approach to Detect Anomalies in Industrial Environments based on Machine Learning, ICPS 2020: 250-256.

[5] Pedro H. J. Nardelli, Constantinos B. Papadias, Charalampos Kalalas, Hirley Alves, Ioannis T. Christou, Irene Macaluso, Nicola Marchetti, Raúl Palacios, Jesus Alonso-Zarate: Framework for the Identification of Rare Events via Machine Learning and IoT Networks, ISWCS 2019: 656-660.

[6] Ioannis T. Christou: Avoiding the Hay for the Needle in the Stack: Online Rule Pruning in Rare Events Detection, ISWCS 2019: 661-665.

[7] I.T. Christou, E. Vagianou, G. Vardoulias, “Planning Courses for Student Success at The American College of Greece”, INFORMS Journal on Applied Analytics, 2023, To Appear.

[8] Routis, G., Paraskevopoulos, M., Vetsikas, I. A., Roussaki, I., Stavrakoudis, D., & Katsantonis, D. (2022, August). Data-Driven and Interoperable Smart Agriculture: An IoT-based Use-Case for Arable Crops. In 2022 IEEE International Conference on Omni-layer Intelligent Systems (COINS). IEEE.

[9] D. Kelesis, D. Vogiatzis, G. Katsimpras, D. Fotakis and G. Paliouras, “Reducing over-smoothing in graph neural networks by changing the activation function,” European Conference on AI (Ecai 2023).

[10] Kefalas, G., & Vogiatzis, D. (2023). Network Structure Versus Chemical Information in Drug-Drug Interaction Prediction. In Complex Networks and Their Applications XI: Proceedings of The Eleventh International Conference on Complex Networks and Their Applications: COMPLEX NETWORKS 2022—Volume 1 (pp. 402-414). Cham: Springer International Publishing.

[11] Vidal ME, Sakor A, Jozashoori S, Niazmand E, Purohit D, Iglesias E, Aisopos F, Vogiatzis D, Menasalvas E, Gonzalez AR, Vigueras G (2023). Knowledge Graphs for Enhancing Transparency in Health Data Ecosystems., Semantic Web Journal

[12] G. Routis, M. Paraskevopoulos, I. A. Vetsikas, I. Roussaki, D. Stavrakoudis, and D. Katsantonis. “Data-Driven and Interoperable Smart Agriculture: An IoT-based Use-Case for Arable Crops.” In 2022 IEEE International Conference on Omni-layer Intelligent Systems (COINS), pp. 1-8. IEEE, 2022.