MiniFool -- Physics-Constraint-Aware Minimizer-Based Adversarial Attacks in Deep Neural Networks
This addresses the problem of evaluating neural network robustness in physics experiments, but it is incremental as it adapts existing adversarial attack concepts to domain-specific constraints.
The authors introduced MiniFool, a physics-constraint-aware adversarial attack algorithm for testing neural networks in particle and astroparticle physics, and applied it to MNIST and CMS Open Data, finding that the likelihood of flipped classifications varies for correctly and incorrectly classified events.
In this paper, we present a new algorithm, MiniFool, that implements physics-inspired adversarial attacks for testing neural network-based classification tasks in particle and astroparticle physics. While we initially developed the algorithm for the search for astrophysical tau neutrinos with the IceCube Neutrino Observatory, we apply it to further data from other science domains, thus demonstrating its general applicability. Here, we apply the algorithm to the well-known MNIST data set and furthermore, to Open Data data from the CMS experiment at the Large Hadron Collider. The algorithm is based on minimizing a cost function that combines a $χ^2$ based test-statistic with the deviation from the desired target score. The test statistic quantifies the probability of the perturbations applied to the data based on the experimental uncertainties. For our studied use cases, we find that the likelihood of a flipped classification differs for both the initially correctly and incorrectly classified events. When testing changes of the classifications as a function of an attack parameter that scales the experimental uncertainties, the robustness of the network decision can be quantified. Furthermore, this allows testing the robustness of the classification of unlabeled experimental data.