Departments of Mathematical Sciences: Workshop on Survival Analysis



  • Time: 9:30am - 11:00am
  • Date: Saturday, June 1st, 2019
  • Venue: MB237

Talk 1


Censored Cumulative Residual Independent Screening for Ultrahigh-Dimensional Survival Data


For complete ultrahigh-dimensional data, sure independent screening methods can effectively reduce the dimensionality while ensuring that all the active variables can be retained with high probability. However, limited screening methods have been developed for censored data, which often arise in clinical trials and genetic studies. We propose a censored cumulative residual independent screening method that is specially tailored to the ultrahigh-dimensional survival data. The proposed screening method is model-free, and it tends to rank the active variables over the inactive ones in terms of their association with the survival times and also enjoys the sure independent screening property. Compared with several existing methods, our model-free screening method works well with general survival models, is invariant to the monotone transformation of the responses, and requires substantially weaker moment conditions. Numerical studies demonstrate the usefulness of the censored cumulative residual independent screening method, and the new approach is illustrated with a gene expression data set.


Yanyan Liu, Professor from the Department of Probability and Statistics, Wuhan University. She is the Executive Member (常务理事) of Chinese Association for Applied Statistics (中国现场统计研究会) and 中国工业统计学会. She is currently the Associate Editor of Statistical papers. Her research interest includes Survival Analysis, Missing Data Analysis and High-Dimension Statistics.

Talk 2


Semiparametric Inference for the Functional Cox Model


We study penalized semiparametric maximum partial likelihood estimation and hypothesis testing for the functional Cox model in analyzing right-censored data with both functional and scalar predictors. Deriving the asymptotic joint distribution of finite-dimensional and infinite-dimensional estimators is a very challenging theoretical problem due to the complexity of semiparametric models. For the problem, we construct the Sobolev space equipped with a special inner product and discover a new joint Bahadur representation of estimators of unknown regression function and coefficients. Using this key tool, we establish the asymptotic properties of the proposed estimators and develop a penalized partial likelihood ratio test. The theoretical results are examined through simulation studies, and a right-censored data example from the Improving Care of Acute Lung Injury Patients study is provided for illustration.


Xingqiu Zhao, Associate Professor from the Department of Applied Mathematics, The Hong Kong Polytechnic University. She is currently the Associate Editor of Journal of Applied Statistics, Communications in Statistics-Theory and Methods, Communications in Statistics-Simulation and Computation, Statistical Methodology, and so on. She has many publications on Annals of Statistics, Bernoulli, and Journal of the American Statistical Association. Her research interest includes Survival Analysis, Panel Count Data, Longitudinal Data Analysis, Recurrent Event Data, Analysis of High-Dimensional Survival Models, Semiparametric and Nonparametric Methods, Large Deviations and Moderate Deviations Theory with Applications in Survival Models and Statistical Machine Learning.