Experimental analysis of feature selection stability for high-dimension and low-sample size gene expression classification task.
David DernoncourtBlaise HanczarJean-Daniel ZuckerPublished in: BIBE (2012)
Keyphrases
- high dimension
- sample size
- small sample
- expression classification
- feature selection
- model selection
- facial expression recognition
- variance reduction
- upper bound
- random sampling
- gene expression data
- real valued
- progressive sampling
- hyperparameters
- microarray
- gene expression
- small samples
- feature space
- worst case
- facial expressions
- support vector machine
- human identification
- input space
- least squares
- computer vision
- unlabeled data
- support vector
- pattern recognition
- expression recognition
- lower bound
- feature vectors
- classification accuracy