Regression with right-censored high-dimensional data: An application with different imputation techniques

This study aims to introduce four modified linear estimators for the right-censored high-dimensional data. Obviously, data of interest involves two important problems to be solved that are censorship and high dimensionality. This paper can be distinguished from other studies in the literature with that it achieves to handle these two problems simultaneously. The main contribution of the paper is merging weightedridge method with the imputation techniques to obtain more efficient estimators than its alternatives. To solve the censorship problem, four imputation techniques are considered based on machine learning algorithms kNN, sliding-windows, regression and support vector machines. The high-dimensionality problem is handled by the weighted ridge approach which provides estimator with less risk than its alternatives because it detects the covariates with a weak contribution via the post-selection procedure. To show the empirical performance of the introduced estimators, a simulation study is made and comparative results are presented. Results show that kNN and regression imputation basis WR esitmators show satisfying performances on estimation of the high-dimensional right-censored model.