On locally optimal tests for the mean direction of the Langevin distribution
Hayakawa (1990) has very recently studied the behavior of the power for several large sample tests for the mean direction vector of the Langevin distribution. These tests are not known to possess any non-trivial optimal property. Here we derive some multiparameter locally optimal tests, e.g., best first and second directional derivative tests and locally most mean powerful tests. For the case when ?, the concentration parameter, is known, these tests are exact and some cut-off points are presented. For the case when ? is unknown, we propose C([alpha])-type tests which are expected to be asymptotically locally optimal. Some open problems are indicated.
Year of publication: |
1991
|
---|---|
Authors: | SenGupta, A. ; Rao Jammalamadaka, S. |
Published in: |
Statistics & Probability Letters. - Elsevier, ISSN 0167-7152. - Vol. 12.1991, 6, p. 537-544
|
Publisher: |
Elsevier |
Saved in:
Online Resource
Saved in favorites
Similar items by person
-
On the limiting distribution of the number of 'near-matches'
Rao Jammalamadaka, S., (1984)
-
Predictive inference for directional data
Rao Jammalamadaka, S., (1998)
-
Large sample distribution of the sample total in a generalized rejective sampling scheme
Rao Jammalamadaka, S., (1991)
- More ...