Random forests with random projections of the output space for high dimensional multi-label classification.
Arnaud JolyPierre GeurtsLouis WehenkelPublished in: CoRR (2014)
Keyphrases
- random forests
- high dimensional
- input space
- dimensionality reduction
- multi label
- low dimensional
- dimension reduction
- high dimensional data
- ensemble methods
- text categorization
- machine learning algorithms
- high dimensionality
- feature space
- decision trees
- class labels
- data points
- graph cuts
- image annotation
- hyperplane
- binary classifiers
- binary classification
- image classification
- unsupervised learning
- learning tasks
- principal component analysis
- training samples
- text classification
- input data
- feature extraction
- kernel function
- information retrieval
- multi class
- pairwise
- training set
- bayesian networks