报告题目:High-dimensional robust inference for censored linear models
报告摘要：Due to the directly statistical interpretation, censored linear regression offers a valuable complement to the Cox proportional hazards regression in survival analysis. Existing work on high-dimensional inference often requires stringent assumptions, such as the independence of censoring time between survival time and covariate or the moment condition of model error, perhaps making some practical applications restrictive. We propose a rank-based high-dimensional inference for censored linear regression. Its validity is guaranteed under a more general and realistic censoring mechanism that survival and censoring times are conditionally independent given covariate and without imposing any moment condition on the model error. We develop theory of high-dimensional $U$-statistic, circumvent challenges stemming from the non-smoothness of loss function, and establish convergence rate of regularized estimator and asymptotic normality of the resulting de-biased estimator as well as consistency of the asymptotic variance estimation. As censoring can be viewed as a manner of trimming, it thereby strengthens the robustness of our proposed rank-based high-dimensional inference, particularly for heavy-tailed model error or outlier in the presence of response. We evaluate the finite-sample performance of the proposed method via extensive simulation studies and demonstrate its utility by applying it to a subcohort study from The Cancer Genome Atlas (TCGA).
报告地点：腾讯会议（581 312 792）