Handling Imprecise Labels in Feature Selection with Graph Laplacian.
Gauthier DoquireMichel VerleysenPublished in: ICPRAM (1) (2012)
Keyphrases
- graph laplacian
- web image annotation
- feature selection
- spectral clustering
- random walk
- label information
- spectral analysis
- dimensionality reduction
- semi supervised learning
- basis functions
- kernel machines
- weighted graph
- euclidean space
- text categorization
- pointwise
- support vector
- pairwise
- manifold structure
- class labels
- labeled and unlabeled data
- semi supervised
- multi label
- label propagation
- labeled data
- text classification
- feature set
- multi class
- unsupervised learning
- support vector machine
- data points
- active learning
- feature extraction
- training data
- neighborhood graph
- machine learning
- manifold learning
- data clustering
- training examples
- supervised learning
- k means
- feature space