DIRECT: Deep Active Learning under Imbalance and Label Noise.
Shyam NuggehalliJifan ZhangLalit JainRobert D. NowakPublished in: CoRR (2023)
Keyphrases
- label noise
- active learning
- class imbalance
- machine learning
- class distribution
- random sampling
- training examples
- unlabeled data
- noise tolerant
- semi supervised
- training set
- supervised learning
- learning algorithm
- labeled data
- cost sensitive
- learning process
- imbalanced datasets
- training data
- visual features
- relevance feedback
- class labels
- binary classification
- sample complexity