Sufficient dimension reduction via distance covariance with multivariate responses

Xianyan Chen, Qingcong Yuan, Xiangrong YinJournal of Nonparametric Statistics 2018 · DOI

Abstract

In this article, we propose a new method for sufficient dimension reduction when both response and predictor are vectors. The new method, using distance covariance, keeps the model-free advantage, and can fully recover the central subspace even when many predictors are discrete. We then extend this method to the dual central subspace, including a special case of canonical correlation analysis. We illustrated estimators through extensive simulations and real datasets, and compared to some existing methods, showing that our estimators are competitive and robust.
BibTeX
@article{chen2018sufficient,
  author = {Xianyan Chen and Qingcong Yuan and Xiangrong Yin},
  title = {{Sufficient dimension reduction via distance covariance with multivariate responses}},
  journal = {Journal of Nonparametric Statistics},
  year = {2018},
  doi = {10.1080/10485252.2018.1562065},
}

Loading related content...

Figure