Non-Asymptotic Oracle Inequalities for the High-Dimensional Cox Regression via Lasso.

Non-Asymptotic Oracle Inequalities for the High-Dimensional Cox Regression via Lasso.
复制标题

DOI:
10.5705/ss.2012.240
复制
发表时间:
2014-01-01
期刊:
影响因子:
1.4
通讯作者:
Nan B
Nan B
中科院分区:
数学3区
文献类型:
--
作者:
Kong S;Nan B

文献摘要

参考文献

被引文献

相似文献

利用套索方法研究了正则化高维Cox回归的有限样本性质。现有文献主要研究具有Lipschitz损失函数的线性模型或广义线性模型,其中经验风险函数是独立同分布损失的总和。然而,对于截尾生存数据,负对数偏似然函数的求和既不是iid也不是Lipschitz,我们首先用iid非Lipschitz项之和来逼近负对数偏似然函数,然后利用逐点论据推导出套索惩罚Cox回归的非渐近预言不等式,以解决因缺乏iid Lipschitz损失而造成的困难.
We consider finite sample properties of the regularized high-dimensional Cox regression via lasso. Existing literature focuses on linear models or generalized linear models with Lipschitz loss functions, where the empirical risk functions are the summations of independent and identically distributed (iid) losses. The summands in the negative log partial likelihood function for censored survival data, however, are neither iid nor Lipschitz.We first approximate the negative log partial likelihood function by a sum of iid non-Lipschitz terms, then derive the non-asymptotic oracle inequalities for the lasso penalized Cox regression using pointwise arguments to tackle the difficulties caused by lacking iid Lipschitz losses.
DOI: 10.1214/08-ejs287
发表时间: 2008-01-01
影响因子: 1.1
作者:
Bunea, Florentina
通讯作者: Bunea, Florentina
DOI: 10.1214/009053606000000092
发表时间: 2006-04-01
影响因子: 4.5
作者:
Buhlmann, Peter
通讯作者: Buhlmann, Peter
DOI: 10.1214/11-aos911
发表时间: 2011
影响因子: 4.5
作者:
Bradic J;Fan J;Jiang J
通讯作者: Jiang J
DOI: 10.1214/09-ejs521
发表时间: 2010-01-01
影响因子: 1.1
作者:
Bach, Francis
通讯作者: Bach, Francis
DOI: 10.1214/009053607000000929
发表时间: 2008-04-01
影响因子: 4.5
作者:
van de Geer, Sara A.
通讯作者: van de Geer, Sara A.