Improvement of Reuse of Classifiers in CBIR Using SVM Active Learning.
Masaaki TekawaMotonobu HattoriPublished in: ICONIP (2) (2010)
Keyphrases
- active learning
- support vector
- training set
- svm classifier
- relevance feedback
- training examples
- training data
- support vector machine classifiers
- classification algorithm
- support vector machine svm
- support vector machine
- improves the classification accuracy
- rule based classifier
- feature selection
- image retrieval
- semi supervised
- decision boundary
- support vectors
- ensemble classifier
- train a support vector machine
- unlabeled instances
- linear support vector machines
- machine learning
- small sample
- classification method
- learning algorithm
- sample selection
- statistical learning theory
- text classifiers
- supervised learning
- kernel support vector machines
- kernel function
- rbf kernel
- classification accuracy
- linear classifiers
- unlabeled data
- active learning framework
- feature vectors
- knn
- hyperplane
- class labels
- training samples
- decision trees
- highest accuracy
- kernel methods
- visual features
- image content
- selective sampling
- data sets
- machine learning algorithms
- feature space
- learning process
- cbir systems
- discriminative classifiers
- linear svm
- svm classification
- feature ranking
- fold cross validation
- image classification
- high classification accuracy
- image representation
- multi class
- rare class
- multiple kernel learning
- multi class classification
- random sampling
- semi supervised learning