Matan Ben-Tov

CaFA: Cost-aware, Feasible Attacks With Database Constraints Against Neural Tabular Classifiers

2024

This work presents CaFA, a system for Cost-aware Feasible Attacks for assessing the robustness of neural tabular classifiers against adversarial examples realizable in the problem space, while minimizing adversaries’ effort. To this end, CaFA leverages …