arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2107.00643cs.LG
收藏

基于分布偏移的Mandoline模型评价

Mandoline: Model Evaluation under Distribution Shift

Stanford University(斯坦福大学)

Mayee Chen, Karan Goel, Nimit S. Sohoni, Fait Poms, Kayvon Fatahalian, Christopher Ré

摘要:

机器学习模型通常部署在不同的环境中,而不是在不同的环境中进行训练和验证,这对希望预测部署的模型在目标分布上的表现的从业者构成了挑战。如果来自目标分布的未标记样本以及来自可能不同的源分布的标记样本可用,则可以应用诸如重要性加权的标准方法来估计目标的性能。然而,当源和目标分布具有不重叠的支持或高维时,重要性加权很难实现。从流行病学和民意调查等领域获得灵感,我们开发了Mandoline,一种新的评估框架,可以缓解这些问题。我们的关键见解是,从业者可能有关于分布偏移的方式的先验知识,我们可以用它来更好地指导重要性加权过程。具体来说,用户编写简单的“切片函数”--有噪声的、潜在相关的二进制函数,旨在捕捉分布偏移的可能轴--来计算重新加权的性能估计。我们进一步描述了一个密度比估计框架的切片,并显示其估计误差尺度与切片质量和数据集大小。NLP和视觉任务的经验验证表明,Mandoline可以比标准基线更准确地估计目标分布的性能高达3倍。

英文摘要:

Machine learning models are often deployed in different settings than they were trained and validated on, posing a challenge to practitioners who wish to predict how well the deployed model will perform on a target distribution. If an unlabeled sample from the target distribution is available, along with a labeled sample from a possibly different source distribution, standard approaches such as importance weighting can be applied to estimate performance on the target. However, importance weighting struggles when the source and target distributions have non-overlapping support or are high-dimensional. Taking inspiration from fields such as epidemiology and polling, we develop Mandoline, a new evaluation framework that mitigates these issues. Our key insight is that practitioners may have prior knowledge about the ways in which the distribution shifts, which we can use to better guide the importance weighting procedure. Specifically, users write simple "slicing functions" - noisy, potentially correlated binary functions intended to capture possible axes of distribution shift - to compute reweighted performance estimates. We further describe a density ratio estimation framework for the slices and show how its estimation error scales with slice quality and dataset size. Empirical validation on NLP and vision tasks shows that Mandoline can estimate performance on the target distribution up to 3x more accurately compared to standard baselines.

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。

↑