|
| << previous | next >> |
Journal description
All volumes of the corresponding journal
All issues of the corresponding volume
All articles of the corresponding issues
Details for article 9 of 26 found articles
| Title: |
BIAS in linear regression models with unknown covariance matrix |
| Author: |
Aubin, Elisete da Conceicao Q. Cordeiro, Gauss M. |
| Appeared in: |
Communications in statistics |
| Paging: | Volume 26 (1997) nr. 3 pages 813-828 |
| Year: |
1997 |
| Contents: |
We investigate the second-order biases of the maximum likelihood estimates from normal linear regression models with unknown error covariance matrix. The error covariance matrix depends on a set of unknown parameters that can be efficiently estimated by maximum likelihood. We give a matrix formula for the n1 biases of the maximum likelihood estimates of these parameters, where n is the sample size. The formula is simple enough to be used algebraically to obtain several closed-form expressions in special cases. It has also advantages for numerical purposes. |
| Publisher: |
Taylor & Francis |
| Source file: |
Elektronische Wetenschappelijke Tijdschriften |
Details for article 9 of 26 found articles
| << previous | next >> |