CFE 2019: Start Registration
View Submission - CMStatistics
Title: Dynamic tilted current correlation for high dimensional variable screening Authors:  Wenqing He - University of Western Ontario (Canada) [presenting]
Abstract: Variable screening is an essential procedure in high dimensional data analysis to reduce dimensionality and ensure the applicability of certain statistical methods. It is a complicated and computationally burdensome procedure since spurious correlations commonly exist among predictor variables, and important predictor variables may not have large marginal correlations with the response variable. We propose a new estimator for the correlation between the response and high-dimensional predictor variables and develop a new variable screening technique for high dimensional data based on the proposed correlation estimator. The proposed screening method enjoys the approximate sure screening and consistency properties and is capable of picking up the relevant predictor variables within a finite number of steps. It has been justified theoretically numerical through simulation and a real gene expression data.