A recursive relation for the inverse of the covariance matrix in optimal pattern classifiers (Corresp).
N. Thomas GaarderPublished in: IEEE Trans. Inf. Theory (1965)
Keyphrases
- covariance matrix
- correlation matrix
- covariance matrices
- class conditional densities
- mahalanobis distance
- estimation error
- training data
- principal component analysis
- sample size
- eigenvalues and eigenvectors
- eigendecomposition
- geometrical interpretation
- positive definite
- objective function
- gaussian mixture
- training samples
- machine learning
- cma es
- symmetric matrix
- pseudo inverse
- principal components
- optimal solution
- evolutionary algorithm
- exhaustive search
- worst case
- feature set