Abstract
This paper investigates the inferential properties of testing the means of Gaussian functional data, using a Mahalanobis type distance for Hilbert spaces. We establish the analytic power of exact and asymptotic tests, for the known and unknown covariance case, respectively.
Lingua originale | English |
---|---|
pagine (da-a) | 102-107 |
Numero di pagine | 6 |
Rivista | STATISTICS & PROBABILITY LETTERS |
Volume | 131 |
DOI | |
Stato di pubblicazione | Pubblicato - 2017 |
Pubblicato esternamente | Sì |
Keywords
- Distances in L2
- Functional data
- Gaussian processes
- Inference on the mean
- Power of exact tests