论文部分内容阅读
For analyzing correlated binary data with high-dimensional covariates,we,in this paper,propose a two-stage shrinkage approach.First,we construct a weighted least-squares(WLS) type function using a special weighting scheme on the non-conservative vector fi