Deep models trade off minimum margin (accuracy) against average margin (robustness); a new regularizer promotes average margin while preserving Fisher consistency.
Making Machine Learning Robust Against Adversarial Inputs.Communications of the ACM, 61 (7):56–66
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Understanding Adversarial Robustness: The Trade-off between Minimum and Average Margin
Deep models trade off minimum margin (accuracy) against average margin (robustness); a new regularizer promotes average margin while preserving Fisher consistency.