Bias of Error Rates in Linear Discriminant Analysis Caused by Feature Selection and Sample Size.
Helene SchulerudPublished in: ICPR (2000)
Keyphrases
- sample size
- small sample size
- variance reduction
- linear discriminant analysis
- feature selection
- small sample
- model selection
- generalization error
- dimensionality reduction
- small samples
- class separability
- discriminant analysis
- feature extraction
- support vector
- dimension reduction
- feature space
- dealing with high dimensional data
- random sampling
- upper bound
- high dimensionality
- linear discriminant
- feature set
- progressive sampling
- principal component analysis
- covariance matrix
- microarray data
- text classification
- face recognition
- worst case
- machine learning
- null space
- support vector machine svm
- knn
- scatter matrices
- classification accuracy
- support vector machine
- pattern recognition
- unsupervised feature selection
- high dimensional data
- computer vision
- inductive learning
- nearest neighbor