Distributed Maximization of "Submodular plus Diversity" Functions for Multi-label Feature Selection on Huge Datasets.
Mehrdad GhadiriMark SchmidtPublished in: CoRR (2019)
Keyphrases
- semi supervised learning
- text categorization
- multi label
- multi label classification
- single label classification
- unlabeled data
- feature selection
- labeled data
- multi label learning
- multi instance
- machine learning
- text classification
- hierarchical text categorization
- objective function
- binary classification
- feature selections
- multiple labels
- image annotation
- learning tasks
- class labels
- naive bayes
- model selection
- graph cuts
- max margin
- data sets
- feature selection algorithms
- multi class
- protein function prediction
- k nearest neighbor
- cost sensitive
- image classification
- nearest neighbor
- label assignment
- feature extraction
- computer vision
- neural network