Department of Statistics, UCLA

A supersaturated design is a design whose run size is not enough for estimating all the main effects. It is commonly used in screening experiment, where the goal is to identify sparse and dominant active effects with low cost. In this paper, we study a variable selection method via Dantzig selector, proposed by Candes and Tao (2007), to screen active effects. A graphical procedure and an automated procedure are suggested to accompany with the method. Simulation studies show that this method is effective over the existing data analysis methods in the literature.