Improving feature extraction by replacing the Fisher criterion by an upper error bound.
Karsten LuebkeClaus WeihsPublished in: Pattern Recognit. (2005)
Keyphrases
- error bounds
- fisher criterion
- discriminant analysis
- feature extraction
- linear discriminant analysis
- fisher linear discriminant
- theoretical analysis
- linear discriminant
- face recognition
- worst case
- dimension reduction
- distance measure
- subspace learning
- dimensionality reduction
- principal component analysis
- feature vectors
- feature selection
- support vector machine svm
- support vector
- machine learning
- high dimensional data
- probability distribution
- feature space
- small sample size
- similarity measure