让机器学习模型在大规模数据上自动实现公平性校准，无需人工指定分组。
MCGrad: Multicalibration at Web Scale
- 自动发现数据子群体进行多校准，无需人工指定保护组。
- 在Meta生产环境中部署，提升模型性能指标如对数损失和PRAUC。
- 算法可扩展，已集成到数百个线上模型，开源可用。
我们提出MCGrad，一种新颖且可扩展的多校准算法。多校准——在数据子群体中实现校准——是机器学习系统性能的重要属性。现有方法因需手动指定子群体、难以扩展或损害对数损失与精确率-召回率曲线下面积（PRAUC）等指标而未被广泛采用。MCGrad无需显式定义受保护群体，具备可扩展性，并通常改善其他性能指标而非损害它们。该算法已在Meta投入生产，现已成为数百个生产模型的一部分。本文展示了实际部署结果及公共数据集上的表现，并在https://github.com/facebookincubator/MCGrad提供开源实现。
原文摘要 · Abstract (English)
We propose MCGrad, a novel and scalable multicalibration algorithm. Multicalibration - calibration in subgroups of the data - is an important property for the performance of machine learning-based systems. Existing multicalibration methods have thus far received limited traction in industry. We argue that this is because existing methods (1) require such subgroups to be manually specified, which ML practitioners often struggle with, (2) are not scalable, or (3) may harm other notions of model performance such as log loss and Area Under the Precision-Recall Curve (PRAUC). MCGrad does not require explicit specification of protected groups, is scalable, and often improves other ML evaluation metrics instead of harming them. MCGrad has been in production at Meta, and is now part of hundreds of production models. We present results from these deployments as well as results on public datasets. We provide an open source implementation of MCGrad at https://github.com/facebookincubator/MCGrad.
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。