New properties for Tyler's covariance matrix estimator.
Gordana DraskovicFrédéric PascalPublished in: ACSSC (2016)
Keyphrases
- covariance matrix
- covariance matrices
- estimation error
- sample size
- principal component analysis
- maximum likelihood
- positive definite
- least squares
- geometrical interpretation
- eigenvalues and eigenvectors
- mahalanobis distance
- pseudo inverse
- objective function
- correlation matrix
- eigendecomposition
- principal components
- cma es
- gaussian mixture
- multivariate gaussian
- face recognition
- class conditional densities
- multivariate normal