Hessian Semi-Supervised Sparse Feature Selection Based on ${L_{2, 1/2}}$ -Matrix Norm.
Caijuan ShiQiuqi RuanGaoyun AnRuizhen ZhaoPublished in: IEEE Trans. Multim. (2015)
Keyphrases
- semi supervised
- low rank
- low rank matrices
- feature selection
- robust principal component analysis
- low rank approximation
- group lasso
- trace norm
- low rank matrix
- norm regularization
- matrix completion
- rank minimization
- subspace learning
- semi supervised learning
- sparse pca
- coefficient matrix
- sparse matrix
- unsupervised learning
- text categorization
- singular value decomposition
- multi task
- supervised learning
- pairwise
- labeled data
- low rank and sparse
- web image annotation
- unlabeled data
- linear combination
- active learning
- kernel matrix
- multi view
- kernel learning
- machine learning
- multi class
- sparse regression
- dimensionality reduction
- partially labeled
- text classification
- feature space
- structured sparsity
- jacobian matrix
- mixed norm
- convex optimization
- feature selection algorithms
- regularized least squares
- matrix factorization
- singular values
- support vector machine
- eigenvalue decomposition
- model selection
- high dimensional data
- missing data
- signal recovery
- classification accuracy
- tensor decomposition
- pairwise constraints
- data matrix
- knn
- feature extraction
- support vector
- high dimensional
- learning tasks
- unsupervised feature selection
- sparse representation
- feature set