Variable Selection Of High-dimensional Nonparameteric Accelerated Failure Time Additive Regression Model

In this article,we will focus on the problem of variable selection in the high-dimensional nonparametric accelerated failure time model.With the development of technology and society,data is becoming more and more complex.Relative to the sample size,the dimensionality of the data will be greater than or far greater than the sample size.Usually we assume that the sample size is n,p is the size of the data and the dimensionality of the data.In this article,we focus on the situation of p>n.However,due to the explosive storage of data,the data collected in some fields are not complete data.Perhaps for some reasons,complete data cannot be observed,that is deleted.The results or responses of the data are censored,which brings a challenge to the development of high-dimensional censored data dimensionality reduction technology,so the research on high-dimensional censored data is particularly important.First,in this article,we briefly outline the development status of the nonparametric accelerated failure time(i.e AFT)model at home and abroad.Secondly,we study the variable selection and estimation problem of the high-dimensional nonparametric accelerated failure time additive regression model.For this problem,we give the estimation method and the estimation method based on the group MCP penalty in the high-dimensional nonparametric accelerated failure time model.It's theoretical properties and proposed algorithm.Finally,we performed Monte Carlo simulations to evaluate the performance of the group MCP penalty method under limited samples.And through actual data analysis to illustrate the usefulness of the proposed method.