提出新方法防止数据组估值被恶意拆分套利
Faithful Group Shapley Value
- 基于数学洞察设计抗攻击的分组数据估值算法
- 计算效率与精度均显著优于现有方法
- 适合需要公平评估数据贡献的协作场景
数据Shapley是衡量单个数据点对机器学习模型贡献的重要工具。实际应用中，当数据提供方以批次形式贡献数据时，需要进行分组层面的数据估值。然而，我们发现现有分组扩展方法易受‘壳公司攻击’影响，即通过策略性拆分数据组可不正当地抬高估值。为此，我们提出忠实分组Shapley值（FGSV），能唯一防御此类攻击。基于原始数学洞察，我们开发了可证明快速且准确的近似算法来计算FGSV。实验表明，该算法在计算效率和近似精度上显著优于当前最优方法，同时确保分组估值的忠实性。
原文摘要 · Abstract (English)
Data Shapley is an important tool for data valuation, which quantifies the contribution of individual data points to machine learning models. In practice, group-level data valuation is desirable when data providers contribute data in batch. However, we identify that existing group-level extensions of Data Shapley are vulnerable to shell company attacks, where strategic group splitting can unfairly inflate valuations. We propose Faithful Group Shapley Value (FGSV) that uniquely defends against such attacks. Building on original mathematical insights, we develop a provably fast and accurate approximation algorithm for computing FGSV. Empirical experiments demonstrate that our algorithm significantly outperforms state-of-the-art methods in computational efficiency and approximation accuracy, while ensuring faithful group-level valuation.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。