Marija Kopanja

Marija Kopanja is a Research Associate and a PhD student in Computer Science at the Faculty of Sciences, University of Novi Sad, where she previously obtained her BSc and MSc degrees in Applied Mathematics. She holds the ECMI Certificate, a diploma supplement awarded by the European Consortium for Mathematics in Industry (ECMI). In 2023, she received the Best Student Research Paper Award at the Faculty of Sciences. She received her master’s degree with a thesis on cost-sensitive learning methods for imbalanced classification problems, which she further extended during her PhD studies by developing a novel machine learning method: Cost-sensitive Rule and Tree Extraction (CORTEX). In addition to machine learning, her research interest is focused on explainable artificial intelligence with applications in various domains, including agriculture. She was a visiting researcher at the School of Computer Science at TU Dublin, where she worked on the development of a new explainable artificial method. She has also been a visiting researcher several times at Wageningen University & Research (WUR), where she worked on applying XAI methods to machine learning models for crop yield prediction using satellite imagery. In collaboration with WUR researchers, she also worked on solutions for creating digital twins for crops by integrating data assimilation techniques with the WOFOST crop simulation model. She is the author and co-author of numerous conference papers, including several journal papers. She has participated in several EU Horizon projects, including ANTARES, UDENE, and DRAGON. She is also actively involved in an industry collaboration project financed by the World Bank.
Center:
CIT
Themes:
1. Machine Learning
2. Deep Learning
3. Knowledge discovery
Projects:
Publications:
- Kopanja, M., Savić, M., Longo, L. (2025). CORTEX: Cost-sensitive rule and tree extraction method. Knowledge-Based Systems. 330. 114592. IF 7.6 10.1016/j.knosys.2025.114592
- Kopanja, M., Savić, M., Longo, L. (2025). Enhancing Cost-Sensitive Tree-Based XAI Surrogate Method: Exploring Alternative Cost Matrix Formulation. The 3rd World Conference on eXplainable Artificial Intelligence: July 09–11, 2025, Istanbul, Turkey
- Kopanja, M., Hačko, S., Brdar, S., Savić, M. (2024) Cost-sensitive tree SHAP for explaining cost-sensitive tree-based models. Computational Intelligence. 2024; 40(3):e12651. doi: 10.1111/coin.12651
- Kopanja, M., Maglevannaia, P., Carević Tomić, M. & Obrenović, N. (2025). Integrating ML and XAI with Urban Planning: Air Quality Predictions to Support Traffic Optimization. 9th International Conference on Advances in Artificial Intelligence (ICAAI 2025), November 14–16, 2025, Manchester, United Kingdom. 10.1145/3787279.3787322
- Van Evert, F. K., Boersma, S., van Oort, P., Maestrini, B.,Kopanja, M., Mimić,G., Pronk, A. (2022). A Digital Twin for Arable Crops and for Grass. Proceedings of the 16th International Conference on Precision Agriculture, Manhattan, USA. The International Society of Precision Agriculture.