Unsupervised graph-based feature selection via subspace and pagerank centrality.
Khadidja HenniNeila MezghaniCharles Gouin-VallerandPublished in: Expert Syst. Appl. (2018)
Keyphrases
- feature selection
- semi supervised
- semi supervised dimensionality reduction
- dimensionality reduction
- unsupervised feature selection
- unsupervised learning
- feature space
- pointwise mutual information
- word sense induction
- supervised and unsupervised learning
- feature extraction
- random walk
- social network analysis
- feature set
- feature subspace
- subspace learning
- high dimensionality
- web search
- text categorization
- low dimensional
- graph connectivity
- information gain
- multi class
- machine learning
- high dimensional
- feature subset
- ranking algorithm
- unlabeled data
- mutual information
- social networks
- text classification
- graph model
- link analysis
- betweenness centrality
- irrelevant features
- normalized cut
- discriminative features
- selected features
- web graph
- network analysis
- network structure
- principal component analysis
- feature selection algorithms
- subspace clusters
- support vector
- supervised learning
- model selection
- face recognition
- decision trees