A technique used during custom training to iteratively select the most informative instances for labeling, thus reducing the overall labeling effort for entity extraction.
Active learning is a technique used during the training of classification or entity extraction models, particularly when custom-training a model. It is a method for the model to interactively select the most informative instances for labeling.
The primary benefit of active learning is that it significantly reduces the overall effort required for manually labeling data for training, making the fine-tuning process more efficient.
Sources & References
Explore other Learning Paradigms terms
R
Reinforcement Learning
An ML category involving learning through trial and error to reach an objective.
S
Supervised Learning
An ML approach used when labeled data is available. Entity extraction (NER) falls under this…
U
Unsupervised Learning
An ML approach used when the model is not told what to look for (no…