Abstract
For a general non-Gaussian stationary linear process, quasi-maximum likelihood estimation of a subset of the parameters of the spectral density is considered when the complementary subset is suspected to be superfluous. A preliminary test quasi-maximum likelihood estimator (q-MLE) of parameters is introduced and, in the light of its mean square error, is compared with the restricted and unrestricted q-MLE.
Original language | English |
---|---|
Pages (from-to) | 1580-1587 |
Number of pages | 8 |
Journal | Statistics and Probability Letters |
Volume | 81 |
Issue number | 11 |
DOIs | |
Publication status | Published - 2011 Nov |
Keywords
- ARMA model
- Preliminary test
- Quasi-maximum likelihood estimator
- Spectral density
- Stationary linear process
ASJC Scopus subject areas
- Statistics and Probability
- Statistics, Probability and Uncertainty