Multiple-Instance Learning by Boosting Infinitely Many Shapelet-based Classifiers.
Daiki SuehiroKohei HatanoEiji TakimotoShuji YamamotoKenichi BannaiAkiko TakedaPublished in: CoRR (2018)
Keyphrases
- randomized trees
- multiple instance learning
- multiple instance
- image categorization
- multi class
- instance selection
- supervised learning
- image annotation
- random forests
- semi supervised
- object based image retrieval
- class labels
- weak classifiers
- ensemble learning
- semi supervised learning
- data sets
- diverse density
- training samples
- decision trees
- feature selection
- ensemble methods
- machine learning algorithms
- labeled data
- small number
- pairwise
- training data