Fundamental questions on robustness and accuracy for classical and quantum learning algorithms

TL;DR

This paper explores the fundamental relationship between accuracy and robustness in classical and quantum classifiers, analyzing conditions for trade-offs and scenarios avoiding them through theoretical insights and toy models.

Contribution

It introduces new definitions of robustness and accuracy, and provides a theoretical framework to understand their interplay in noisy and adversarial settings for classical and quantum algorithms.

Findings

Trade-offs between accuracy and robustness can arise under certain conditions.

Some scenarios allow for avoiding the accuracy-robustness trade-off.

The framework highlights the influence of noise, bias, and perturbations on model performance.

Abstract

This chapter introduces and investigates some fundamental questions on the relationship between accuracy and robustness in both classical and quantum classification algorithms under noisy and adversarial conditions. We introduce and clarify various definitions of robustness and accuracy, including corrupted-instance robustness accuracy and prediction-change robustness, distinguishing them from conventional accuracy and robustness measures. Through theoretical analysis and toy models, we establish conditions under which trade-offs between accuracy and robustness accuracy arise and identify scenarios where such trade-offs can be avoided. The framework developed highlights the nuanced interplay between model bias, noise characteristics, and perturbation types, including relevant and irrelevant perturbations. We explore the implications of some of these results for incompatible noise, adversarial quantum perturbations, the no free lunch theorem, and suggest future methods to examine these problems from the lens of dynamical systems.