Adversarial Attacks, Regression, and Numerical Stability Regularization

TL;DR

This paper investigates adversarial attacks on regression neural networks, proposing a regularization-based defense that enhances numerical stability and outperforms previous methods.

Contribution

It introduces a novel regularization technique to defend against adversarial attacks in regression, focusing on improving numerical stability of learned functions.

Findings

The proposed regularization outperforms prior defenses.

The method improves numerical stability of regression models.

Adversarial attacks are linked to numerical instability.

Abstract

Adversarial attacks against neural networks in a regression setting are a critical yet understudied problem. In this work, we advance the state of the art by investigating adversarial attacks against regression networks and by formulating a more effective defense against these attacks. In particular, we take the perspective that adversarial attacks are likely caused by numerical instability in learned functions. We introduce a stability inducing, regularization based defense against adversarial attacks in the regression setting. Our new and easy to implement defense is shown to outperform prior approaches and to improve the numerical stability of learned functions.