Volume 2,Issue 7
基于半参数变换模型的右删失生存数据纵向联邦统计学习
处于大数据时代的今天，分布式存储是医疗领域常见的生存数据存储模式，这给传统的生存分析模型带来了新的挑战。基于此，本文基于半参数变换模型研究右删失生存数据的纵向联邦统计学习问题。首先，基于全部样本构建全局似然函数，通过引入各机构本地特征线性组合中间变量，将全局似然函数转化为各机构可联合优化的似然函数；然后，提出了一套相应的ADMM 优化算法，允许各机构与中心服务器之间进行隐私数据信息传输，同时保持了集中式极大似然估计的效率；最后统计大量模拟例子和实例分析说明了所研究方法的有效性和合理性。
[1]Cox DR. Regression models and life-tables[J]. Journal of the Royal Statistical Society: Series B (Methodological), 1972, 34(2): 187-202.
[2]Cox DR. Partial likelihood[J]. Biometrika, 1975, 62(2): 269-276.
[3]Ren JJ, Zhou M. Full likelihood inferences in the Cox model: an empirical likelihood approach[J]. Annals of the Institute of Statistical Mathematics, 2011, 63(5): 1005-1018.
[4]Cheng SC, Wei LJ, Ying Z. Analysis of transformation models with censored data[J]. Biometrika, 1995, 82(4): 835-845.
[5]Zeng D, Lin YD. Maximum likelihood estimation in semiparametric regression models with censored data[J]. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 2007, 69(4): 507-564.
[6]Achcar JA, Barili E, Martinez EZ. Survival analysis of critically ill patients with cancer: use of semiparametric (transformation models) under a hierarchical Bayesian
approach[J]. Brazilian Journal of Biometrics, 2025, 43(3): e-43722.
[7]Lu CL, Wang S, Ji Z, et al. WebDISCO: a web service for distributed cox model learning without patient-level data sharing[J]. Journal of the American Medical Informatics Association, 2015, 22(6): 1212-1219.
[8]Duan R, Boland MR, Liu Z, et al. Learning from electronic health records across multiple sites: A communication-efficient and privacy-preserving distributed algorithm[J].Journal of the American Medical Informatics Association, 2020, 27(3): 430-437.
[9]Dai W, Jiang X, Bonomi L, et al. VERTICOX: Vertically distributed Cox proportional hazards model using the alternating direction method of multipliers[J]. IEEE
Transactions on Knowledge and Data Engineering, 2022, 34(2): 996-1010.