A Maximum Uncertainty LDA-Based Approach for Limited Sample Size Problems : With Application to Face Recognition.
Carlos E. ThomazDuncan Fyfe GilliesPublished in: SIBGRAPI (2005)
Keyphrases
- sample size
- small sample size
- high dimensionality
- pac learning
- random sampling
- statistical tests
- discriminant analysis
- linear discriminant analysis
- upper bound
- microarray data
- face recognition
- number of training samples
- covariance matrix
- model selection
- dimensionality reduction
- small samples
- confidence intervals
- experimental design
- vc dimension
- small sample
- optimality conditions
- scatter matrices
- variance reduction
- data sets
- null space
- inductive learning
- topic models
- active learning
- computational complexity