Training robust and generalizable quantum models

arXiv:2311.11871v2 Announce Type: replace-cross
Abstract: Adversarial robustness and generalization are both crucial properties of reliable machine learning models. In this letter, we study these properties in the context of quantum machine learning based on Lipschitz bounds. We derive parameter-dependent Lipschitz bounds for quantum models with trainable encoding, showing that the norm of the data encoding has a crucial impact on the robustness against data perturbations. Further, we derive a bound on the generalization error which explicitly involves the parameters of …

abstract adversarial arxiv context cs.lg data encoding machine machine learning machine learning models math.oc norm quant-ph quantum robust robustness study training type