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 2409.14700cs.CR
收藏

用于生成表格数据的自适应鲁棒水印

Adaptive and Robust Watermark for Generative Tabular Data

JPMorganChase(摩根大通)

Dung Daniel Ngo, Archan Ray, Akshay Seshadri, Daniel Scott, Saheed Obitayo, Niraj Kumar, Vamsi K. Potluru, Marco Pistoia, Manuela Veloso

AI总结:

研究为防止生成表格数据被滥用的水印算法,提出一种灵活鲁棒的算法,给出其在多指标上性能的理论保证,证明技术有独立价值,通过合成和真实数据集验证了理论结果。

摘要:

近年来,为防止合成数据被滥用,给生成表格数据添加水印成为一个重要框架。多数现有表格数据水印方法虽有多种优良特性,但结果多为针对常见不经意和对抗性攻击的经验保证。本文研究一种灵活且鲁棒的生成表格数据水印算法,给出该算法在保真度、可检测性、鲁棒性和解码难度等指标上性能的理论保证,其证明技术有独立价值且可用于机器学习其他领域,最后在合成和真实世界表格数据集上验证了理论结果。

英文摘要:

In recent years, watermarking generative tabular data has become a prominent framework to protect against the misuse of synthetic data. However, while most prior work in watermarking methods for tabular data demonstrate a wide variety of desirable properties (e.g., high fidelity, detectability, robustness), the findings often emphasize empirical guarantees against common oblivious and adversarial attacks. In this paper, we study a flexible and robust watermarking algorithm for generative tabular data. Specifically, we demonstrate theoretical guarantees on the performance of the algorithm on metrics like fidelity, detectability, robustness, and hardness of decoding. The proof techniques introduced in this work may be of independent interest and may find applicability in other areas of machine learning. Finally, we validate our theoretical findings on synthetic and real-world tabular datasets.

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。

↑