Adaptive Adversarial Training Does Not Increase Recourse Costs

Abstract

Recent work has connected adversarial attack methods and algorithmic recourse methods: both seek minimal changes to an input instance which alter a model's classification decision. It has been shown that traditional adversarial training, which seeks to minimize a classifier's susceptibility to malicious perturbations, increases the cost of generated recourse; with larger adversarial training radii correlating with higher recourse costs. From the perspective of algorithmic recourse, however, the appropriate adversarial training radius has always been unknown. Another recent line of work has motivated adversarial training with adaptive training radii to address the issue of instance-wise variable adversarial vulnerability, showing success in domains with unknown attack radii. This work studies the effects of adaptive adversarial training on algorithmic recourse costs. We establish that the improvements in model robustness induced by adaptive adversarial training show little effect on algorithmic recourse costs, providing a potential avenue for affordable robustness in domains where recoursability is critical.

References in corpus (10)

• Theoretically Principled Trade-off between Robustness and Accuracy
• Actionable Recourse in Linear Classification
• Adversarial Attacks and Defences: A Survey
• Learning Model-Agnostic Counterfactual Explanations for Tabular Data
• Instance adaptive adversarial training: Improved accuracy tradeoffs in neural nets
• Counterfactual Explanations and Algorithmic Recourses for Machine Learning: A Review
• Imperceptible Adversarial Attacks on Tabular Data
• Recent Advances in Adversarial Training for Adversarial Robustness
• A Framework and Benchmarking Study for Counterfactual Generating Methods on Tabular Data
• CAT: Customized Adversarial Training for Improved Robustness