A general solution to the maximization of the multidimensional generalized Rayleigh quotient used in linear discriminant analysis for signal classification.
Ramon PrietoPublished in: ICASSP (6) (2003)
Keyphrases
- linear discriminant analysis
- discriminant analysis
- principal component analysis
- face recognition
- dimensionality reduction
- feature extraction
- support vector machine svm
- linear discriminant
- feature space
- high dimensional data
- dimension reduction
- null space
- subspace methods
- principal components analysis
- discriminant features
- support vector
- small sample size
- fisher criterion
- scatter matrix
- qr decomposition
- multivariate statistical
- finite dimensional
- discriminative information
- nearest neighbor
- pattern recognition
- nonparametric discriminant analysis