A Feasible Nonconvex Relaxation Approach to Feature Selection.
Cuixia GaoNaiyan WangQi Rose YuZhihua ZhangPublished in: AAAI (2011)
Keyphrases
- feature selection
- objective function
- text categorization
- feature set
- feasible solution
- selected features
- mutual information
- multi class
- iterative algorithms
- machine learning
- high dimensionality
- model selection
- classification accuracy
- probabilistic relaxation
- support vector machine
- information gain
- convex optimization
- nonlinear programming
- unsupervised feature selection
- feature selection algorithms
- multi task
- classification models
- feature subset
- method for feature selection
- feature extraction
- text classification
- convex functions
- small sample
- quadratic optimization problems
- global optimization
- optimization problems
- cost function
- lower bound
- bayesian networks
- similarity measure
- image segmentation
- learning algorithm