Feature combination using linear discriminant analysis and its pitfalls.
Ralf SchlüterAndrás ZolnayHermann NeyPublished in: INTERSPEECH (2006)
Keyphrases
- linear discriminant analysis
- discriminant features
- discriminant analysis
- principal component analysis
- dimensionality reduction
- face recognition
- feature analysis
- locality preserving projections
- feature extraction
- small sample size
- null space
- dimension reduction
- high dimensional data
- support vector
- principal components analysis
- fisher criterion
- discriminative information
- linear discriminant
- support vector machine svm
- feature space
- subspace methods
- scatter matrix
- feature vectors
- class separability
- multivariate statistical
- high dimensionality
- scatter matrices
- intra class
- data sets
- covariance matrix
- svm classifier
- sample size
- kernel function
- low dimensional
- signal processing
- feature set
- image features
- data points
- supervised dimensionality reduction
- qr decomposition
- pattern recognition
- dealing with high dimensional data