Back to Search
Start Over
Quasi shrinkage estimation of a block-structured covariance matrix.
- Source :
-
Journal of Statistical Computation & Simulation . Sep2024, Vol. 94 Issue 14, p3093-3110. 18p. - Publication Year :
- 2024
-
Abstract
- In this paper, we study the estimation of a block covariance matrix with linearly structured off-diagonal blocks. We consider estimation based on the least squares method, which has some drawbacks. These estimates are not always well conditioned and may not even be definite. We propose a new estimation procedure providing a structured positive definite and well-conditioned estimator with good statistical properties. The least squares estimator is improved with the use of a shrinkage method and an additional algebraic approach. The resulting so-called quasi shrinkage estimator is compared with the structured maximum likelihood estimator. [ABSTRACT FROM AUTHOR]
Details
- Language :
- English
- ISSN :
- 00949655
- Volume :
- 94
- Issue :
- 14
- Database :
- Academic Search Index
- Journal :
- Journal of Statistical Computation & Simulation
- Publication Type :
- Academic Journal
- Accession number :
- 179995565
- Full Text :
- https://doi.org/10.1080/00949655.2024.2367164