Authorship Verification by combining SVMs with Kernels Optimized for Different Feature Categories.
Julián Solórzano-SotoVictor MijangosAlejandro PimentelFernanda López-EscobedoAzucena MontesGerardo SierraPublished in: CLEF (Working Notes) (2015)
Keyphrases
- support vector
- kernel function
- svm classification
- kernel methods
- feature vectors
- feature maps
- feature space
- multiple kernel learning
- margin maximization
- feature selection
- support vector machine
- input features
- additional features
- polynomial kernels
- feature ranking
- broad categories
- decision trees
- image features
- kernel machines
- rbf kernel
- multiple kernel
- large margin classifiers
- gaussian kernel
- sparse kernel
- support vector machine svm
- learning machines
- support vectors
- feature extraction
- knn
- kernel fisher discriminant analysis
- training data
- category specific
- training set
- classification using support vector machines
- kernel parameters
- single feature
- loss function
- linear svm
- multi class
- semi supervised
- training samples
- generalization ability