Robust nonparametric regression on Riemannian Manifolds

GUILLERMO HENRY; DANIELA RODRIGUEZ

Vigo, Espana

Congreso; International Seminar on Nonparametric Inference; 2008

Universidad de Vigo

In this paper, we propose two families of robust kernel based regression estimators when the regressors are random objects taking values in a Riemannian manifold. We consider estimators based on kernel methods and estimators based on k-nearest neighbor kernel methods. Strong uniform consistency results as well as the asymptotically normality of both families are established.