Semi-supervised SVMs for classification with unknown class proportions and a small labeled dataset.
Sathiya Keerthi SelvarajBigyan BharSundararajan SellamanickamShirish K. ShevadePublished in: CIKM (2011)
Keyphrases
- semi supervised
- support vector
- supervised learning
- multi class classification
- multiclass classification
- unlabeled samples
- support vector machine svm
- training set
- label information
- feature selection
- class labels
- support vector machine
- machine learning
- multi class
- svm classifier
- classification algorithm
- labeled data
- benchmark datasets
- classification method
- classification accuracy
- feature extraction
- feature vectors
- partially labeled
- cost sensitive
- training data
- fold cross validation
- data generating distribution
- support vector machine classifiers
- decision functions
- unsupervised learning
- feature space
- labeled examples
- decision boundary
- text classification tasks
- training samples
- semi supervised classification
- feature set
- manually labeled
- decision trees
- text classification
- training support vector machines
- category labels
- multiclass problems
- imbalanced datasets
- kernel machines
- active learning
- generalization ability
- training dataset
- kernel function
- labeled and unlabeled data
- object classes