SUN’IY INTELLEKT TIZIMLARIDA O‘QITILGAN MODELNI HIMOYALASH ARXITEKTURALARI
Keywords:
Sun’iy intellekt, o‘qitilgan model, xavfsizlik, modelni himoyalash, ko‘p qismli xavfsiz hisoblash, gomomorfik shifrlash, federativ o‘qitish, differensial maxfiylik, suv belgisiga asoslangan modelAbstract
Sun'iy intellekt (SI) sohasining jadal rivojlanishi bilan o‘qitilgan modellarni himoyalash masalasi tobora dolzarb muammoga aylanmoqda. Ushbu modellar katta hajmdagi ma'lumotlar va yuqori hisoblash resurslari asosida yaratilgan qimmatli intellektual mulk sifatida ruxsatsiz foydalanish, ko‘chirish yoki hujumlardan himoyalanishni talab qiladi. Mazkur maqolada SI tizimlarida o‘qitilgan modellarni himoyalash uchun qo‘llaniladigan asosiy arxitekturalar tahlil qilinadi. Tadqiqotning maqsadi ushbu usullarning samaradorligini va SI ilovalarining o‘ziga xos talablariga muvofiqligini baholashdir.
References
Zhou, F. Tofigh, M. Piccardi, M. Abolhasan, D. Franklin and J. Lipman, "Secure Multi-Party Computation for Machine Learning: A Survey," in IEEE Access, vol. 12, pp. 53881-53899, 2024, doi: 10.1109/ACCESS.2024.3388992.
R. Podschwadt, D. Takabi, P. Hu, M. H. Rafiei and Z. Cai, "A Survey of Deep Learning Architectures for Privacy-Preserving Machine Learning With Fully Homomorphic Encryption," in IEEE Access, vol. 10, pp. 117477-117500, 2022, doi: 10.1109/ACCESS.2022.3219049.
Yong Cheng, Yang Liu, Tianjian Chen, and Qiang Yang. 2020. Federated learning for privacy-preserving AI. Commun. ACM 63, 12 (December 2020), 33–36. https://doi.org/10.1145/3387107
Jiang, Honglu & Gao, Yifeng & Sarwar, S M & GarzaPerez, Luis & Robin, Mahmudul. (2022). Differential Privacy in Privacy-Preserving Big Data and Learning: Challenge and Opportunity. 10.1007/978-3-030-96057-5_3.
Tong Qiao, Yuyan Ma, Ning Zheng, Hanzhou Wu, Yanli Chen, Ming Xu, Xiangyang Luo. A novel model watermarking for protecting generative adversarial network. Computers & Security. Volume 127, 2023, 103102, ISSN 0167-4048, https://doi.org/10.1016/j.cose.2023.103102.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2025 Bozorov Suhrobjon Mumin o‘g‘li
This work is licensed under a Creative Commons Attribution 4.0 International License.
