Mirza,F.K.Ogrenci,A.S.2024-06-232024-06-2320230979-835034294-9https://doi.org/10.1109/CINTI59972.2023.10382025https://hdl.handle.net/20.500.12469/5813Using various methods of computational intelligence, scores for credit applications are predicted for making a decision to accept or to reject. Past data about credits accepted in a financial institution are used to develop hybrid models implementing gradient boosting and attention supported neural networks. The performance (Gini scores above 0.7), limitations and further research directions are discussed. © 2023 IEEE.eninfo:eu-repo/semantics/closedAccessattention mechanismcredit scoringgradient boostingneural networksunbalanced dataConference Object11111610.1109/CINTI59972.2023.103820252-s2.0-85184121699