A kernel support vector machine-based feature selection approach for recognizing Flying Apsaras' streamers in the Dunhuang Grotto Murals, China.
Zhong ChenShengwu XiongZhixiang FangQingquan LiBaolin WangQin ZouPublished in: Pattern Recognit. Lett. (2014)
Keyphrases
- feature selection
- support vector
- class separability
- feature space
- kernel function
- kernel learning
- kernel methods
- text classification
- multiple kernel learning
- feature set
- classification accuracy
- text categorization
- feature extraction
- support vector machine
- automatic recognition
- feature subset
- unsupervised feature selection
- feature selection algorithms
- linear svm
- vision system
- multi class
- knn
- high dimensionality
- similarity function
- information gain
- economic development
- mutual information
- kernel regression
- hilbert schmidt independence criterion
- neural network
- positive definite
- kernel pca
- feature ranking
- multi task
- model selection
- principal component analysis
- machine learning