Timezone: »

Evaluating Robustness of Predictive Uncertainty Estimation: Are Dirichlet-based Models Reliable?
Anna-Kathrin Kopetzki · Bertrand Charpentier · Daniel Zügner · Sandhya Giri · Stephan Günnemann

Wed Jul 21 05:40 AM -- 05:45 AM (PDT) @ None

Dirichlet-based uncertainty (DBU) models are a recent and promising class of uncertainty-aware models. DBU models predict the parameters of a Dirichlet distribution to provide fast, high-quality uncertainty estimates alongside with class predictions. In this work, we present the first large-scale, in-depth study of the robustness of DBU models under adversarial attacks. Our results suggest that uncertainty estimates of DBU models are not robust w.r.t. three important tasks: (1) indicating correctly and wrongly classified samples; (2) detecting adversarial examples; and (3) distinguishing between in-distribution (ID) and out-of-distribution (OOD) data. Additionally, we explore the first approaches to make DBU mod- els more robust. While adversarial training has a minor effect, our median smoothing based ap- proach significantly increases robustness of DBU models.

Author Information

Anna-Kathrin Kopetzki (Technical University of Munich)
Bertrand Charpentier (Technical University of Munich)
Daniel Zügner (Technical University of Munich)

PhD candidate @ **TU Munich**. Research on robust machine learning for graphs. Previous: R&D intern @ **Apple**, machine learning for music data.

Sandhya Giri (Technical University of Munich)
Stephan Günnemann (Technical University of Munich)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors