Subsampling Strategies¶
For large datasets, making predictions for the whole unlabeled set on each iteration to obtain the uncertainty estimates may require an enormous amount of time and resources. Unlabeled pool subsampling algorithms are adressing this issue by subsampling instances in the unlabeled pool depending on their uncertainty scores obtained on previous AL iterations. This helps to speed up the AL iterations, especially when the unlabeled pool is large.
# |
Strategy |
Citation |
1 |
||
2 |
||
3 |
- |