Unified improvements in estimation of a normal covariance matrix in high and low dimensions.
Hisayuki TsukumaTatsuya KubokawaPublished in: J. Multivar. Anal. (2016)
Keyphrases
- covariance matrix
- estimation error
- covariance matrices
- sample size
- principal component analysis
- positive definite
- multivariate gaussian
- gaussian mixture
- mahalanobis distance
- objective function
- multivariate normal
- geometrical interpretation
- correlation matrix
- pseudo inverse
- cma es
- eigenvalues and eigenvectors
- special case
- maximum likelihood estimation
- closed form
- multi objective
- image processing