Stochastic approximation of the factors of a generalized canonical correlation analysis
We suppose that data of a generalized canonical correlation analysis are i.i.d. observations of a random vector Z which are taken sequentially. We define a recursive method of sequential estimation of the factors. This can be applied also when there are a great number of non random data vectors.
Year of publication: |
2008
|
---|---|
Authors: | Monnez, Jean-Marie |
Published in: |
Statistics & Probability Letters. - Elsevier, ISSN 0167-7152. - Vol. 78.2008, 14, p. 2210-2216
|
Publisher: |
Elsevier |
Saved in:
Online Resource
Saved in favorites
Similar items by person
-
A fast and recursive algorithm for clustering large datasets with k-medians
Cardot, Hervé, (2012)
-
Almost sure convergence of stochastic gradient processes with matrix step sizes
Monnez, Jean-Marie, (2006)
- More ...