摘要

This paper aims at developing a powerful step-up procedure controlling the generalized family-wise error rate (k-FWER) in multiple hypotheses testing, based on the least favorable configuration of k-FWER. Simulation studies suggest improvement over some existing methods.

全文