arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2004.05675cs.LGstat.ML
收藏

生成模型中数据不一致性的非参数检验

A Non-Parametric Test to Detect Data-Copying in Generative Models

University of California, San Diego(加州大学圣地亚哥分校)

Casey Meehan, Kamalika Chaudhuri, Sanjoy Dasgupta

摘要:

在生成模型中检测过拟合是机器学习中的一个重要挑战。在这项工作中,我们形式化了一种过拟合的形式,我们称之为{ \em{data-copying} } -其中生成模型存储并输出训练样本或其小的变化。我们提供了一个用于检测数据复制的三样本非参数测试,该测试使用训练集、来自目标分布的单独样本和来自模型的生成样本,并研究了我们的测试在几个规范模型和数据集上的性能。有关代码\和示例,请访问https://github.com/casey-meehan/data-copying

英文摘要:

Detecting overfitting in generative models is an important challenge in machine learning. In this work, we formalize a form of overfitting that we call {\em{data-copying}} -- where the generative model memorizes and outputs training samples or small variations thereof. We provide a three sample non-parametric test for detecting data-copying that uses the training set, a separate sample from the target distribution, and a generated sample from the model, and study the performance of our test on several canonical models and datasets. For code \& examples, visit https://github.com/casey-meehan/data-copying

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。

↑