Steganalysis in high dimensions: fusing classifiers built on random subspaces.
Jan KodovskýJessica J. FridrichPublished in: Media Forensics and Security (2011)
Keyphrases
- high dimensions
- random subspaces
- co training
- high dimensional data
- high dimensional
- decision trees
- training data
- poor quality
- class imbalance
- semi supervised learning
- high dimensional spaces
- nearest neighbor classifiers
- test set
- training examples
- support vector
- data mining
- training set
- feature set
- unlabeled data
- ensemble classifier
- machine learning methods
- multiple classifier systems
- feature subspace
- classifier fusion
- semi supervised