Principal Component Analysis in Hilbert Space: a Perturbation Approach.
The setup of our study is the Principal Component Analysis for dependent Hilbert space valued random vectors with examples taken from the ARH (1) model. We prove that convergence rates of the j ^(th) empirical random projection operator II from n to j for almost sure convergence and convergence in distribution depend on rates of convergence of the aasociate sequence of covariance operators.
Year of publication: |
1999
|
---|---|
Authors: | Mas, A. |
Subject: | MATHEMATICAL ANALYSIS | ECONOMETRICS |
Saved in:
Saved in favorites
Similar items by subject
-
Croux, C., (1999)
-
Orthogonality of the Sheffer Polynomials Associated to a Levy Process.
Pommeret, D., (1999)
-
The Unbiased Estimation Theory and the Multidimensional Bhattachryya Bounds.
Pommeret, D., (1999)
- More ...
Similar items by person