arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2105.12697cs.LGcs.CR
收藏

结构因果模型揭示线性规划建模中的混杂偏差

Structural Causal Models Reveal Confounder Bias in Linear Program Modelling

Technical University of Darmstadt(达姆施塔特工业大学);Eindhoven University of Technology(埃因霍温理工大学);DFKI(德国人工智能研究中心)

Matej Zečević, Devendra Singh Dhami, Kristian Kersting

AI总结:

本研究借助结构因果模型揭示线性规划中的混杂偏差,证明其可引发对抗式攻击,并为三类组合问题提供存在性证明,拓展了对抗攻击的研究范畴。

摘要:

近年来,针对对抗攻击的研究不断拓展,尤其是在深度神经网络领域。在本研究中,我们提出并探究一个问题:该现象在本质上是否更具普遍性,即经典分类任务之外是否存在对抗式攻击。具体而言,我们研究了作为现代人工智能研究基础组成部分的优化问题。为此,我们选取了优化器的基础类别——线性规划(Linear Programs, LPs)。在最初尝试将对抗样本的形式体系与线性规划进行简单映射时,我们很快发现,要为线性规划构建合理的对抗样本概念,还缺少关键要素。有趣的是,Pearl的因果概念形式体系能够恰当地描述线性规划的类对抗样本。我们的核心结论表明,结构因果模型(Structural Causal Model, SCM)对后续的线性规划优化有直接影响,这最终揭示了线性规划中的混杂概念(由上述SCM引入),而这种混杂使得对抗式攻击成为可能。我们既提供了形式化的通用证明,也针对三个组合问题——线性分配、最短路径以及一个能源系统的实际问题,给出了基于SCM的这类特殊LP参数化的存在性证明。

英文摘要:

The recent years have been marked by extended research on adversarial attacks, especially on deep neural networks. With this work we intend on posing and investigating the question of whether the phenomenon might be more general in nature, that is, adversarial-style attacks outside classical classification tasks. Specifically, we investigate optimization problems as they constitute a fundamental part of modern AI research. To this end, we consider the base class of optimizers namely Linear Programs (LPs). On our initial attempt of a naïve mapping between the formalism of adversarial examples and LPs, we quickly identify the key ingredients missing for making sense of a reasonable notion of adversarial examples for LPs. Intriguingly, the formalism of Pearl's notion to causality allows for the right description of adversarial like examples for LPs. Characteristically, we show the direct influence of the Structural Causal Model (SCM) onto the subsequent LP optimization, which ultimately exposes a notion of confounding in LPs (inherited by said SCM) that allows for adversarial-style attacks. We provide both the general proof formally alongside existential proofs of such intriguing LP-parameterizations based on SCM for three combinatorial problems, namely Linear Assignment, Shortest Path and a real world problem of energy systems.

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。

↑