Reversing uncertainty sampling to improve active learning schemes
Active learning provides promising methods to optimize the cost of manually annotating a dataset. However, practitioners in many areas do not massively resort to such methods because they present technical difficulties and do not provide a guarantee of good performance, especially in skewed distribu...
Guardado en:
| Autores principales: | , |
|---|---|
| Formato: | Objeto de conferencia |
| Lenguaje: | Inglés |
| Publicado: |
2015
|
| Materias: | |
| Acceso en línea: | http://sedici.unlp.edu.ar/handle/10915/52131 http://44jaiio.sadio.org.ar/sites/default/files/asai184-191.pdf |
| Aporte de: |
| Sumario: | Active learning provides promising methods to optimize the cost of manually annotating a dataset. However, practitioners in many areas do not massively resort to such methods because they present technical difficulties and do not provide a guarantee of good performance, especially in skewed distributions with scarcely populated minority classes and an undefined, catch-all majority class, which are very common in human-related phenomena like natural language.
In this paper we present a comparison of the simplest active learning technique, pool-based uncertainty sampling, and its opposite, which we call reversed uncertainty sampling. We show that both obtain results comparable to the random, arguing for a more insightful approach to active learning. |
|---|