OSP-Class: Open Set Pseudo-labeling with Noise Robust Training for Text Classification.
Dohyung KimJahwan KooUng-Mo KimPublished in: Big Data (2022)
Keyphrases
- text classification
- image noise
- text classifiers
- noisy environments
- text mining
- greater robustness
- text categorization
- training corpus
- noisy data
- training examples
- training samples
- linear svm
- naive bayes
- supervised learning
- salt pepper
- active learning
- training set
- training data
- image segmentation
- geometric distortions
- text classification tasks
- training and testing data
- learning algorithm
- unsupervised learning
- feature selection
- imaging artifacts
- labeled data for training
- data cleaning
- signal to noise ratio
- bag of words
- n gram
- labeled data