VisActive: Visual-concept-based Active Learning for Image Classification under Class Imbalance.
Mohammed KhaleelAzeez IdrisWallapak TavanapongJacob PrattJung-Hwan OhPiet C. de GroenPublished in: ACM Trans. Multim. Comput. Commun. Appl. (2024)
Keyphrases
- class imbalance
- active learning
- image classification
- cost sensitive learning
- cost sensitive
- sampling methods
- learning algorithm
- class distribution
- supervised learning
- imbalanced class distribution
- imbalanced datasets
- imbalanced data classification
- majority class
- feature extraction
- machine learning
- minority class
- unlabeled data
- small disjuncts
- training set
- misclassification costs
- semi supervised
- concept drift
- random sampling
- multi label
- imbalanced data
- data analysis
- training examples
- labeled data
- multi class
- image features
- high dimensionality
- high dimensional
- error rate
- learning environment
- text categorization