右删失数据下多响应AFT模型的两阶段估计
A Two-Stage Estimation of Multiple-Response AFT Model with Right-Censored Data
-
摘要: 在生存分析领域,加速失效时间(AFT)模型经常被用于预测事件发生的时间.本文将该模型推广到多事件时间情形, 提出了多响应AFT模型,并假设协变量是高维的, 模型的系数矩阵是联合低秩且稀疏的此外还假设多个事件时间受制于同一个右删失变量.为了估计模型中的系数矩阵, 本文提出一个两阶段方法,先对数据进行逆概率删失加权(IPCW), 再用SESS算法求解一个稀疏降秩回归问题.本文通过数值模拟, 验证了所提方法的有效性.最后将该方法应用于一个关于白血病患者骨髓移植的临床数据集.Abstract: In survival studies, the accelerated failure time (AFT) model is often applied to predict the event times. This article proposes a multiple-response AFT model that extends the AFT model to the multiple events case. It is assumed that the covariates are high-dimensional and the regression coefficient matrix is jointly low-rank and sparse. We also assume all the multivariate event times are subject to right-censoring by a common censoring variable. To estimate the coefficient matrix, a two-stage procedure is proposed. First weight the data with IPCW weights, and then use SESS algorithm to solve a sparse reduced-rank regression problem. The simulation results show that the proposed method performs well in many cases. The method is also applied to a real dataset of bone marrow transplant patients.