This study investigates the asymptotic distribution of linear combinations of parameter estimators in directed weighted network model under differential privacy constraints. It is demonstrated that as the number of network vertices increases, the linear combination of model parameter estimators converge to an asymptotic normal distribution, revealing its asymptotic theory under differential privacy protection. Furthermore, numerical simulations validate the theoretical effectiveness, providing novel theoretical tools and analytical methodologies for statistical inference of network data under differential privacy protection.
DWORKC, MCSHERRYF, NISSIMK, et al. Calibrating noise to sensitivity in private data analysis[C]//ACM SIGSAC.Theory of Cryptography: Third Theory of Cryptography Conference. New York: Springer Berlin Heidelberg, 2006: 265-284.
[2]
LUW, MIKLAUG. Exponential random graph estimation under differential privacy[C]//ACM SIGKDD. 20th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. London: ACM, 2014: 921-930.
[3]
KARWAV, SLAVKOVIĆA. Inference using noisy degrees: Differentially private β-model and synthetic graphs[J]. Ann Statist, 2016, 44(1): 87-112.
[4]
PANL, YANT. Asymptotics in the β-model for networks with a differentially private degree sequence[J]. Communications in Statistics-Theory and Methods, 2020, 49(18): 4378-4393.
[5]
LUOJ, QINH. Asymptotic in a class of network models with a difference private degree sequence[J]. Statistics and Its Interface, 2022, 15(3): 383-397.
[6]
YANT. Directed networks with a differentially private bi-degree sequence[J]. Statistica Sinica, 2021, 31(4): 2031-2050.
[7]
WANGQ, ZHANGX, LUOJ, et al. Weighted directed networks with a differentially private bi-degree sequence[J]. Communications in Statistics-Theory and Methods, 2022, 51(2): 285-300.
[8]
LUOJ, LIUT, WANGQ. Affiliation weighted networks with a differentially private degree sequence[J]. Statistical Papers, 2022, 63(2): 367-395.
[9]
LUOJ, QINH. Asymptotic in the ordered networks with a noisy degree sequence[J]. Journal of Systems Science and Complexity, 2022, 35(3): 1137-1153.
[10]
WANGQ, YANT, JIANGB, et al. Two-mode networks: Inference with as many parameters as actors and differential privacy[J]. Journal of Machine Learning Research, 2022, 23(292): 1-38.
[11]
YANT. Differentially private analysis of networks with covariates via a generalized β-model[J]. arXiv:2023,2311.10279.
[12]
PANL, HUJ, LIP. Asymptotic theory in a class of directed random graph models with a differentially private bi-degree sequence[J]. arXiv:2022,2201.09648.
[13]
OUYANGY, JINGL, QIUPINGW, et al. Asymptotics in the Bradley-Terry model for networks with a differentially private degree sequence[J]. Communications in Statistics-Theory and Methods, 2025, 54(2): 437-456.
[14]
PANL, HUJ. Differentially private estimation in a class of bipartite graph models[J]. Communications in Statistics-Theory and Methods, 2024, 53(18): 6477-6496.
[15]
SALASJ, GONZÁLEZ-ZELAYAV, TORRAV, et al. Differentially private graph publishing through noise-graph addition[C]//Springer Nature Switzerland. International Conference on Modeling Decisions for Artificial Intelligence. Cham: Springer Nature Switzerland, 2023: 253-264.
LUOJ, QINH, WANGZ. Asymptotic distribution in directed finite weighted random graphs with an increasing bi-degree sequence[J]. Acta Mathematica Scientia, 2020, 40(2): 355-368.
[19]
LUOJ, QINH, YANT, et al. A note on asymptotic distributions in directed exponential random graph models with bi-degree sequences[J]. Communications in Statistics-Theory and Methods,2017,46(18): 8852-8864.
[20]
YANT, LENGC, ZHUJ. Asymptotics in directed exponential random graph models with an increasing bi-degree sequence[J]. The Annals of Statistics,2016,44(1):31-57.
[21]
BILLINGSLEYP. Convergence of probability measures[M]. New York: John Wiley & Sons, 2013.