arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1903.01032cs.LGcs.SYeess.SYstat.ML
收藏

对抗分类中的基本性能限制

A Fundamental Performance Limitation for Adversarial Classification

Abed AlRahman Al Makdah, Vaibhav Katewa, Fabio Pasqualetti

AI总结:

本文研究了对抗分类中的基本性能限制,指出在优化准确率的过程中,二分类算法不可避免地会变得更加敏感于数据的对抗操纵,并且准确率与敏感度之间的根本权衡曲线仅取决于数据的统计特性,无法通过调整算法来改进。

摘要:

尽管机器学习算法被广泛用于解决技术、经济和社会相关的问题,但对这些数据驱动算法性能的可证明保证却严重不足,尤其是在数据来自不可靠来源并通过未保护和易受攻击的通道传输时。在本文中,我们采取了重要的步骤来弥合这一差距,并正式证明,在试图优化其准确率时,二分类算法——包括基于机器学习技术的算法——不可避免地会变得更加敏感于数据的对抗操纵。进一步地,对于具有相同复杂度(即分类边界数量)的给定算法类,准确率与敏感度之间的根本权衡曲线仅取决于数据的统计特性,无法通过调整算法来改进。

英文摘要:

Despite the widespread use of machine learning algorithms to solve problems of technological, economic, and social relevance, provable guarantees on the performance of these data-driven algorithms are critically lacking, especially when the data originates from unreliable sources and is transmitted over unprotected and easily accessible channels. In this paper we take an important step to bridge this gap and formally show that, in a quest to optimize their accuracy, binary classification algorithms -- including those based on machine-learning techniques -- inevitably become more sensitive to adversarial manipulation of the data. Further, for a given class of algorithms with the same complexity (i.e., number of classification boundaries), the fundamental tradeoff curve between accuracy and sensitivity depends solely on the statistics of the data, and cannot be improved by tuning the algorithm.

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。

↑