Raqamli Transformatsiya va Sun’iy Intellekt
JADVALLI MA’LUMOTLARDAGI BO‘SH VA ANOMAL QIYMATLARNI ANIQLASH HAMDA QAYTA TIKLASH ASOSIDA DASTLABKI MA’LUMOTLARNI QAYTA ISHLASH ALGORITMI
1 Muhammad al-Xorazmiy nomidagi Toshkent axborot texnologiyalari universiteti
2 Denov tadbirkorlik va pedagogika instituti
Zenodo DOI: 10.5281/zenodo.23344757
Keywords
Abstract
Ushbu maqolada mashinali o‘qitish masalalarida foydalaniladigan jadvalli ma’lumotlarning sifatini oshirishga qaratilgan adaptiv dastlabki ishlov berish algoritmi tadqiq etilgan. Jadvalli ma’lumotlar tarkibida uchraydigan bo‘sh qiymatlar, noto‘g‘ri kiritilgan hamda predmet sohasi uchun mantiqan yaroqsiz bo‘lgan anomal qiymatlar klassifikatsiya modellarining samaradorligiga salbiy ta’sir ko‘rsatishi mumkin. Mazkur muammoni bartaraf etish maqsadida belgining turi, ruxsat etilgan qiymatlar sohasi va obyektning sinf tegini hisobga oluvchi algoritmik yondashuv taklif etilgan. Algoritmda sonli va tartibli belgilar uchun mediana, nominal va binar belgilar uchun esa moda qiymatidan foydalaniladi. Bunda statistik qiymatlar butun dataset asosida emas, faqat o‘qitish to‘plamidagi to‘g‘ri qiymatlar asosida shakllantiriladi. Bu esa test ma’lumotlarining o‘qitish jarayoniga bilvosita ta’sir ko‘rsatishi, ya’ni ma’lumotlarning sizib chiqishi xavfini kamaytiradi.
References
- Ribeiro C. E., Freitas A. A data-driven missing value imputation approach for longitudinal datasets // Artificial Intelligence Review. – 2021. – Vol. 54, No. 8. – P. 1–31. DOI: 10.1007/s10462-021-09963-5.
- Afkanpour M., Hosseinzadeh E., Tabesh H. Identify the most appropriate imputation method for handling missing values in clinical structured datasets: A systematic review // BMC Medical Research Methodology. – 2024. – Vol. 24, No. 1. – P. 1–13. DOI: 10.1186/s12874-024-02310-6.
- Nishanov A. Kh., Jurayev G. P., Narziyev N. B. Adaptable genetic algorithm for primary processing of medical data based on random search // Shestaya mezhdunarodnaya nauchno-prakticheskaya konferentsiya “BIG DATA and Advanced Analytics”. Ch. 1. – Minsk, 20–21 maya 2020 g. – P. 147–155.
- Nishanov A. Kh., Ollamberganov F. F., Khasanova M. A., Mengturayev F. Z., Allayarov U. B. Algorithms for clustering, classification and selection of informative features for early detection of brain cancer // Acta IMEKO. – 2025. – Vol. 14, No. 3. – P. 1–8. DOI: 10.21014/actaimeko.v14i3.2053
- Hu Y.-H., Wu R.-Y., Lin Y.-C., Lin T.-Y. A novel MissForest-based missing values imputation approach with recursive feature elimination in medical applications // BMC Medical Research Methodology. – 2024. – Vol. 24, No. 1. – P. 1–12. DOI: 10.1186/s12874-024-02392-2.
- Linkon A. A. Evaluation of feature transformation and machine learning models on early detection of diabetes mellitus // IEEE Access. – 2024. – Vol. 12. – P. 165425–165440. DOI: 10.1109/ACCESS.2024.3488743
- Jarrett D., Cebere B. C., Liu T., Curth A., van der Schaar M. HyperImpute: Generalized iterative imputation with automatic model selection // Proceedings of the 39th International Conference on Machine Learning (ICML). – PMLR, 2022. – P. 9916–9937. DOI: 10.48550/arXiv.2206.07769
- Shin K., Han J., Kang S. MI-MOTE: Multiple imputation-based minority oversampling technique for imbalanced and incomplete data classification // Information Sciences. – 2021. – Vol. 575. – P. 80–89. DOI: 10.1016/j.ins.2021.06.043.
- Nishanov A. Kh., Jo‘rayev G‘. P., Azatov B. A. Ma’lumotlarga dastlabki ishlov berishning ayrim masalalarini yechish bosqichlari // V Mezhdunarodnaya nauchno-prakticheskaya konferentsiya “Nauka i obrazovanie v sovremennom mire: vyzovy XXI veka”. Tom I. – Nur-Sultan, 10–12 dekabrya 2019 g. – P. 23–27.
- Hubbard R. A., Lett E., Ho G. Y. F., Chubak J. Characterizing bias due to differential exposure ascertainment in electronic health record data // Health Services and Outcomes Research Methodology. – 2021. – Vol. 21, No. 3. – P. 309–323. DOI: 10.1007/s10742-020-00235-3.
- Alghawli A. S. Complex methods detect anomalies in real time based on time series analysis // Alexandria Engineering Journal. – 2022. – Vol. 61, No. 1. – P. 549–561. DOI: 10.1016/j.aej.2021.06.033.
- Nishonov A.X., Mengturayev F. Z., Ollamberganov F.F., Allayarov U.B., Khasanova M.A., Doniyorova G.T. An algorithm for creating a semi-synthetic dataset for diabetes KazNU Bulletin. Mathematics, Mechanics, Computer Science Series ISSN 1563–0277 Vol. 129 No. 1 (2026): -P. 114-128
- Barrabes M., Perera M., Novelle Moriano V., Giro-I-Nieto X., Mas Montserrat D., Ioannidis A. G. Advances in biomedical missing data imputation: A survey // IEEE Access. – 2025. – Vol. 13. – P. 16918–16932. DOI: 10.1109/ACCESS.2024.3516506.
- Agbo B., Al-Aqrabi H., Alsboui T., Hussain M., Hill R. Imputation of missing clinical covariates for downstream classification problems // IEEE Access. – 2023. – Vol. 11. – P. 102935–102943. DOI: 10.1109/ACCESS.2023.3317775.
- Nishanov A. Kh., Djurayev G. P. Solving stages of classification in the preliminary treating of medical examples // Tekhnicheskie nauki: problemy i resheniya: sbornik statey XXXIII mezhdunarodnoy nauchno-prakticheskoy konferentsii. No. 2 (30). – Moskva, 2020. – P. 69–76.
- Altukhova O. Choice of method imputation missing values for obstetrics clinical data // Procedia Computer Science. – 2020. – Vol. 176. – P. 976–984. DOI: 10.1016/j.procs.2020.09.093.