2021
DOI: 10.48550/arxiv.2110.03802
|View full text |Cite
Preprint
|
Sign up to set email alerts
|

Hitting the Target: Stopping Active Learning at the Cost-Based Optimum

Abstract: Active learning allows machine learning models to be trained using fewer labels while retaining similar performance to traditional fully supervised learning. An active learner selects the most informative data points, requests their labels, and retrains itself. While this approach is promising, it leaves an open problem of how to determine when the model is 'good enough' without the additional labels required for traditional evaluation. In the past, different stopping criteria have been proposed aiming to iden… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1

Citation Types

0
1
0

Year Published

2022
2022
2022
2022

Publication Types

Select...
1

Relationship

0
1

Authors

Journals

citations
Cited by 1 publication
(1 citation statement)
references
References 19 publications
0
1
0
Order By: Relevance
“…A widely applicable stopping method based on stabilizing predictions is presented in [8]. This is a particularly important method to investigate because it is widely applicable with few if any restrictions on when it can be used, and it has been found to work well in many settings [8], [14], [19]- [24]. The method stops when successively trained models have high agreement in terms of their predictions on the examples in a large randomly selected set of examples.…”
Section: Related Workmentioning
confidence: 99%
“…A widely applicable stopping method based on stabilizing predictions is presented in [8]. This is a particularly important method to investigate because it is widely applicable with few if any restrictions on when it can be used, and it has been found to work well in many settings [8], [14], [19]- [24]. The method stops when successively trained models have high agreement in terms of their predictions on the examples in a large randomly selected set of examples.…”
Section: Related Workmentioning
confidence: 99%