2026-06-05T22:04:44.426Z http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai

ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai:3:11959 2026-05-31T00:20:27Z 3

Classification of Hospital Admission Utilizing CatBoost Tuned by Adapted Artificial Bee Colony Metaheuristics, Chapter in LNNS Lecture Notes in Networks and Systems: ICSISCET 2025: Artificial Intelligence and Sustainable Computing, Springer, volume 1938 2026 http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai/record/3/11959 https://link.springer.com/chapter/10.1007/978-3-032-23945-7_23 M. Varsandán J. Petrovic P. Dabic T. Zivkovic M. Zivkovic B. Radomirovic S. Janicijevic N. Bacanin Precise assessment of patient risk at the time of hospital admission is essential for effective prioritization and to guarantee prompt and appropriate medical care. This research focuses on predicting various clinical results using data collected during admission to a cardiac treatment unit, applying an enhanced and fine-tuned machine learning strategy. The categorization task was executed with the CatBoost classifier, whose parameters were carefully optimized through a customized variant of the artificial bee colony metaheuristic technique. The introduced framework achieved notable accuracy in identifying cardiac-related outcomes, with the top models reaching approximately 91.47% correctness. Thus, the suggested approach revealed strong promise as a supportive clinical decision-making instrument, capable of facilitating timely interventions and improving the distribution and utilization of hospital capacities in the future. bookPart Springer, Cham 272 286 10.1007/978-3-032-23945-7_23 LNNS Lecture Notes in Networks and Systems: ICSISCET 2025: Proceedings of International Conference on Sustainable and Innovative Solutions for Current Challenges in Engineering & Technology, volume 1938