Trustworthy Actionable Perturbations

arXiv:2405.11195v1 Announce Type: new
Abstract: Counterfactuals, or modified inputs that lead to a different outcome, are an important tool for understanding the logic used by machine learning classifiers and how to change an undesirable classification. Even if a counterfactual changes a classifier's decision, however, it may not affect the true underlying class probabilities, i.e. the counterfactual may act like an adversarial attack and ``fool'' the classifier. We propose a new framework for creating modified inputs that change the true underlying …

abstract arxiv change class classification classifier classifiers counterfactual cs.ai cs.it cs.lg decision however inputs logic machine machine learning math.it tool true trustworthy type understanding