|
Avtomatika i Telemekhanika, 1983, Issue 2, Pages 167–170
(Mi at5073)
|
|
|
|
Notes
On one problem of robust estimation from correlated observations
N. V. Luneva Minsk
Abstract:
A robust estimate is obstained for a regression function parameter with random dependent noises. The error distribution density is assumed to belong to a distribution class with a constrained covariance matrix. The “worst” distribution which minimizes the Fisher information matrix is proved to be normal distribution. In the case of a linear regression model the resultant estimate is found to be optimal in the minimax sense.
Received: 19.06.1981
Citation:
N. V. Luneva, “On one problem of robust estimation from correlated observations”, Avtomat. i Telemekh., 1983, no. 2, 167–170
Linking options:
https://www.mathnet.ru/eng/at5073 https://www.mathnet.ru/eng/at/y1983/i2/p167
|
|