A threshold method for imbalanced multiple noisy labeling.
Jing ZhangXindong WuVictor S. ShengPublished in: ASONAM (2013)
Keyphrases
- experimental evaluation
- detection method
- preprocessing
- pairwise
- synthetic data
- threshold selection
- high precision
- computational complexity
- data sets
- edge detection
- high accuracy
- test data
- classification method
- segmentation method
- missing data
- support vector machine svm
- input data
- classification accuracy
- probabilistic model
- significant improvement
- prior knowledge
- support vector
- image processing
- neural network