arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2008.01710cs.LGcs.GTstat.ML
收藏

战略感知器

The Strategic Perceptron

University of Maryland(马里兰大学);Northwestern University(西北大学);Toyota Technological Institute at Chicago(芝加哥丰田技术研究所);Amazon(亚马逊)

Saba Ahmadi, Hedyeh Beyhaghi, Avrim Blum, Keziah Naggita

摘要:

经典的感知器算法提供了一个简单而优雅的学习线性分类器的过程。在每一步中,该算法观察样本的位置和标签,并在出错时相应地更新当前预测器。然而,在存在期望被分类为阳性并且能够以有限的量修改其位置的策略性试剂的情况下,分类器可能无法观察试剂的真实位置,而是试剂假装的位置。与具有完美位置知识的原始设置不同,在这种情况下,Perceptron算法无法实现其保证,并且我们说明了预测器永远在两个解决方案之间振荡的示例,即使存在完美的大余量线性分类器,也会产生无限数量的错误。我们的主要贡献是提供了一个修改后的感知器风格的算法,使有限数量的错误存在的战略代理与$\ell_2$和加权$\ell_1$操纵成本。在我们的基线模型中,操纵成本的知识(即,代理可以操纵的程度)。在我们最一般的模型中,我们放松了这一假设,并提供了一种算法,该算法可以学习和改进分类器及其成本估计,即使在操作成本未知的情况下也可以实现良好的错误范围。

英文摘要:

The classical Perceptron algorithm provides a simple and elegant procedure for learning a linear classifier. In each step, the algorithm observes the sample's position and label and updates the current predictor accordingly if it makes a mistake. However, in presence of strategic agents that desire to be classified as positive and that are able to modify their position by a limited amount, the classifier may not be able to observe the true position of agents but rather a position where the agent pretends to be. Unlike the original setting with perfect knowledge of positions, in this situation the Perceptron algorithm fails to achieve its guarantees, and we illustrate examples with the predictor oscillating between two solutions forever, making an unbounded number of mistakes even though a perfect large-margin linear classifier exists. Our main contribution is providing a modified Perceptron-style algorithm which makes a bounded number of mistakes in presence of strategic agents with both $\ell_2$ and weighted $\ell_1$ manipulation costs. In our baseline model, knowledge of the manipulation costs (i.e., the extent to which an agent may manipulate) is assumed. In our most general model, we relax this assumption and provide an algorithm which learns and refines both the classifier and its cost estimates to achieve good mistake bounds even when manipulation costs are unknown.

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。

↑