1 paper
Jiashun Jin, Cun-Hui Zhang, Qi Zhang
Consider a linear regression model where the design matrix X has n rows and p columns. We assume (a) p is much large than n, (b) the coefficient vector beta is sparse in the sense…