Summary form only given, as follows. The problem of estimating the number of 'systematic' components present in the data is reviewed. In this context, a new test based on higher powers of eigenvalues of the sample covariance matrix is derived. The usefulness of this test when the number of systematic components is greater than the dimensionality of the data vector for a certain type of structured covariance matrix is demonstrated.
|Original language||English (US)|
|Number of pages||1|
|State||Published - 1986|
ASJC Scopus subject areas