Unsupervised maximum margin feature selection via L 2, 1-norm minimization.
Shizhun YangChenping HouFeiping NieYi WuPublished in: Neural Comput. Appl. (2012)
Keyphrases
- maximum margin
- unsupervised feature selection
- feature selection
- support vector
- support vector machine
- multi task
- hyperplane
- markov networks
- pattern classification
- unsupervised learning
- learning algorithm
- multiple kernel learning
- semidefinite programming
- feature subset
- machine learning
- high order
- multi class
- convex optimization
- data sets
- classification accuracy
- knn
- sparse coding
- feature set
- text categorization
- maximum likelihood
- graphical models
- feature space
- pattern recognition
- bayesian networks
- face recognition
- interior point methods
- model selection