This is an old revision of this page, as edited by The Anome (talk | contribs) at 12:39, 10 September 2004 (→See also: * Independent components analysis). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.
Revision as of 12:39, 10 September 2004 by The Anome (talk | contribs) (→See also: * Independent components analysis)(diff) ← Previous revision | Latest revision (diff) | Newer revision → (diff)You must add a |reason= parameter to this Cleanup template – replace it with {{Cleanup|reason=<Fill reason here>}}, or remove the Cleanup template.
In statistics, a random vector is said to be "white" if it has the following properties: that the elements are uncorrelated and have unit variance. This corresponds to a flat power spectrum.
A vector can be whitened to remove these correlations. This is useful in various procedures such as data compression.
Whitening a signal
- X_white = E * A' * X
where X is the matrix to be whitened, E is the column matrix of Eigenvectors and A' is the transposed diagonal matrix of eigenvalues.
This article is a stub. You can help Misplaced Pages by expanding it. |