RACT is a rank-adaptive permutation test for two-sample covariance matrices that targets low-rank differences via the Ky-Fan(k) norm to improve power while maintaining exact Type I error control.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ME 1years
2023 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Rank-adaptive covariance testing with applications to genomics and neuroimaging
RACT is a rank-adaptive permutation test for two-sample covariance matrices that targets low-rank differences via the Ky-Fan(k) norm to improve power while maintaining exact Type I error control.