[2603.23318] Robustness Quantification for Discriminative Models: a New Robustness Metric and its Application to Dynamic Classifier Selection

Computer Science > Machine Learning arXiv:2603.23318 (cs) [Submitted on 24 Mar 2026] Title:Robustness Quantification for Discriminative Models: a New Robustness Metric and its Application to Dynamic Classifier Selection Authors:Rodrigo F. L. Lassance, Jasper De Bock View a PDF of the paper titled Robustness Quantification for Discriminative Models: a New Robustness Metric and its Application to Dynamic Classifier Selection, by Rodrigo F. L. Lassance and Jasper De Bock View PDF HTML (experimental) Abstract:Among the different possible strategies for evaluating the reliability of individual predictions of classifiers, robustness quantification stands out as a method that evaluates how much uncertainty a classifier could cope with before changing its prediction. However, its applicability is more limited than some of its alternatives, since it requires the use of generative models and restricts the analyses either to specific model architectures or discrete features. In this work, we propose a new robustness metric applicable to any probabilistic discriminative classifier and any type of features. We demonstrate that this new metric is capable of distinguishing between reliable and unreliable predictions, and use this observation to develop new strategies for dynamic classifier selection. Subjects: Machine Learning (cs.LG); Machine Learning (stat.ML) Cite as: arXiv:2603.23318 [cs.LG]   (or arXiv:2603.23318v1 [cs.LG] for this version)   https://doi.org/10.48550/arXiv.2603.2331...