aufe
Task-Agnostic Undesirable Feature Deactivation Using Out-of-Distribution Data Supplementary Material A Broader Impact
When labeling cost is high, DNNs are inevitably trained in few-shot learning settings. An application whose labeling cost is very high, e.g., For ease of exposition, let's consider a binary classification setting. The other training configurations are the same as those in 4.1.
Country:
- North America > Canada > Ontario > Toronto (0.14)
- North America > United States > California (0.04)
Technology:
- Information Technology > Artificial Intelligence > Vision (1.00)
- Information Technology > Artificial Intelligence > Natural Language (1.00)
- Information Technology > Artificial Intelligence > Machine Learning > Statistical Learning (0.72)
- Information Technology > Artificial Intelligence > Machine Learning > Neural Networks > Deep Learning (0.69)