arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1811.04422cs.LGstat.ML
收藏

对抗性机器学习的最优控制观

An Optimal Control View of Adversarial Machine Learning

University of Wisconsin-Madison(威斯康星大学麦迪逊分校)

Xiaojin Zhu

摘要:

我描述了对抗性机器学习的最优控制视图,其中动态系统是机器学习者,输入是对抗性动作,控制成本由对手的目标定义,以造成伤害并难以检测。这种观点涵盖了许多类型的对抗性机器学习,包括测试项攻击、训练数据中毒和对抗性奖励塑造。该观点鼓励对抗性机器学习研究人员利用控制理论和强化学习的进步。

英文摘要:

I describe an optimal control view of adversarial machine learning, where the dynamical system is the machine learner, the input are adversarial actions, and the control costs are defined by the adversary's goals to do harm and be hard to detect. This view encompasses many types of adversarial machine learning, including test-item attacks, training-data poisoning, and adversarial reward shaping. The view encourages adversarial machine learning researcher to utilize advances in control theory and reinforcement learning.

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。

↑