The moment of inertia and the linear discriminant function
In this note, we show that the characteristic vector of the moment of inertia matrix associated with the first or last characteristic root corresponds to the best linear discriminant function in the situation where the data is a mixture of two multivariate normal distributions with proportional covariance matrices. This result may prove useful as a part of many outlier detection methods. We also describe a small simulation study which illustrates the computational efficiency of the new method.
Year of publication: |
2005
|
---|---|
Authors: | Reyen, Salem S. ; Miller, John J. |
Published in: |
Statistics & Probability Letters. - Elsevier, ISSN 0167-7152. - Vol. 71.2005, 1, p. 39-46
|
Publisher: |
Elsevier |
Keywords: | Linear discriminant analysis Moment of inertia Outlier detection |
Saved in:
Online Resource
Saved in favorites
Similar items by person
-
Economics of alternative coal transportation and distribution systems in Iowa
Baumel, Phillip, (1978)
-
Impact of rail abandonment upon grain elevator and rural community performance measures
Miller, John J., (1977)
-
Devaluation of foreign currencies
Miller, John J., (1967)
- More ...