arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1708.07807cs.CRcs.LGstat.ML
收藏

模块化学习组件攻击:今天的现实,明天的挑战

Modular Learning Component Attacks: Today's Reality, Tomorrow's Challenge

Lehigh University(里海大学)

Xinyang Zhang, Yujie Ji, Ting Wang

摘要:

今天的许多机器学习(ML)系统都不是从头开始构建的,而是由{ \em模块化学习组件}(MLC)阵列组成的。MLCs的使用越来越多,大大简化了ML系统的开发周期。然而,由于大多数多边契约是由第三方提供和维护的,它们缺乏标准化和监管,这带来了深刻的安全影响。在本文中,我们首次证明了潜在有害的MLCs对ML系统的安全性构成了巨大的威胁。我们提出了一个广泛的{ \em逻辑炸弹}攻击,恶意制作的MLC触发主机系统以可预测的方式发生故障。通过实证研究医疗保健领域的两个最先进的ML系统,我们探索了这种攻击的可行性。例如,我们表明,在没有关于主机ML系统的先验知识的情况下,通过仅修改MLC的参数中的3.3{ \textperthousand },每个参数的失真都低于$10^{-3}$,对手能够以100%的成功率迫使误诊目标受害者的皮肤癌。我们为这种攻击的成功提供了分析理由,指出了当今ML模型的基本特征:高维性,非线性和非凸性。因此,这个问题似乎是许多ML系统的基础。我们进一步讨论了潜在的对策,以减轻基于MLC的攻击及其潜在的技术挑战。

英文摘要:

Many of today's machine learning (ML) systems are not built from scratch, but are compositions of an array of {\em modular learning components} (MLCs). The increasing use of MLCs significantly simplifies the ML system development cycles. However, as most MLCs are contributed and maintained by third parties, their lack of standardization and regulation entails profound security implications. In this paper, for the first time, we demonstrate that potentially harmful MLCs pose immense threats to the security of ML systems. We present a broad class of {\em logic-bomb} attacks in which maliciously crafted MLCs trigger host systems to malfunction in a predictable manner. By empirically studying two state-of-the-art ML systems in the healthcare domain, we explore the feasibility of such attacks. For example, we show that, without prior knowledge about the host ML system, by modifying only 3.3{\textperthousand} of the MLC's parameters, each with distortion below $10^{-3}$, the adversary is able to force the misdiagnosis of target victims' skin cancers with 100\% success rate. We provide analytical justification for the success of such attacks, which points to the fundamental characteristics of today's ML models: high dimensionality, non-linearity, and non-convexity. The issue thus seems fundamental to many ML systems. We further discuss potential countermeasures to mitigate MLC-based attacks and their potential technical challenges.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑