Revisit Few-shot Intent Classification with PLMs: Direct Fine-tuning vs. Continual Pre-training.
Haode ZhangHaowen LiangLi-Ming ZhanXiao-Ming WuAlbert Y. S. LamPublished in: CoRR (2023)
Keyphrases
- fine tuning
- training phase
- training set
- supervised learning
- pattern recognition
- classification accuracy
- training process
- training samples
- avoid overfitting
- viable alternative
- image classification
- automatic classification
- fine tune
- fine tuned
- labelled data
- decision trees
- classification systems
- training dataset
- classification scheme
- pattern classification
- classification method
- training procedure
- classification process
- classification performances
- search engine
- classification algorithm
- support vector
- machine learning algorithms
- feature vectors
- support vector machine svm
- visual features
- unsupervised learning
- machine learning
- feature extraction
- video sequences
- training patterns
- feature space
- support vector machine
- model selection
- text categorization
- video data
- training algorithm
- benchmark datasets
- classification models
- test set
- machine learning methods
- decision rules