Learning a kernel function for classification with small training samples.
Tomer HertzAharon Bar-HillelDaphna WeinshallPublished in: ICML (2006)
Keyphrases
- training samples
- feature space
- kernel function
- supervised learning
- learning algorithm
- kernel learning
- decision boundary
- hyperplane
- kernel matrix
- training set
- support vector machine
- active learning
- learning process
- positive training examples
- support vectors
- high dimensional
- input space
- feature vectors
- reinforcement learning
- support vector
- dimensionality reduction
- representative samples
- kernel machines
- training data
- training dataset
- machine learning
- test sample
- learning tasks
- kernel methods
- classification accuracy
- face images
- positive examples
- base classifiers
- discriminative information
- high dimensional feature space
- optimal kernel