Adversarial Training of Linear Models under Stealthy Attacks

2026-08-26

AI Summary

为解决预测模型在面对隐蔽攻击时的脆弱性问题，提出基于检测器的切换模型及线性模型的对抗性风险凸公式化方法。

Abstract

Predictive models are widely used in many fields, but are vulnerable to false data injection attacks. To address this, detection schemes and adversarial training have been proposed, but such approaches lack guarantees against stealthy attacks. We therefore propose a detector-based switched model, in which optimal attack strategies are stealthy. For linear prediction models, we derive a convex formulation of the resulting adversarial risk. The model incorporates protected features and introduces a hyperparameter modelling attack probability, enabling an explicit performance trade-off between clean and attacked data regimes. Numerical simulations on real and synthetic data show improved performance on partially attacked data, even for misspecified attack probabilities.

Problem

• Adversarial Training
• Stealthy Attacks
• Linear Models
• False Data Injection

Innovation

• detector-based switched model
• convex formulation of adversarial risk
• protected features
• hyperparameter for attack probability