Semi-Supervised Novelty Detection Using SVM Entire Solution Path.
Frank de MorsierDevis TuiaMaurice BorgeaudVolker GassJean-Philippe ThiranPublished in: IEEE Trans. Geosci. Remote. Sens. (2013)
Keyphrases
- novelty detection
- solution path
- semi supervised
- support vector
- svm classifier
- support vector machine svm
- loss function
- semi supervised learning
- support vector machine
- anomaly detection
- semi supervised classification
- knn
- piecewise linear
- pairwise
- concept drift
- support vector regression
- labeled data
- convex programming
- unsupervised learning
- elastic net
- training data
- hyperplane
- regularization parameter
- supervised learning
- kernel function
- feature vectors
- pairwise constraints
- support vectors
- feature selection
- classification algorithm
- active learning
- null space
- linear support vector machines
- unlabeled data
- optimal solution
- data sets
- k nearest neighbor
- training set
- machine learning
- kernel methods
- cross validation
- clustering algorithm
- logistic regression
- data streams
- classification accuracy