Definition #
Uses small labeled datasets alongside large unlabeled data to improve accuracy.
Key Characteristics #
- Cost-effective
- Leverages unlabeled data
- Common in NLP
Why It Matters #
Reduces data labeling costs by up to 60%.
Examples #
- Speech recognition systems
- Document classification