数据子群体间机器学习表现非线性相关性（英文版）

英文标题：On the nonlinear correlation of ML performance between data subpopulations中文摘要：研究机器学习模型在不同的数据分布下的性能表现，发现在子人群变化时，性能间的相关性呈现 “月形” 相关性，并且这种非线性相关性受到训练数据中虚假相关的影响，研究结果对机器学习的可靠性和公平性具有应用意义。英文摘要：Understanding the performance of machine learning (ML) models across diversedata distributions is critically important for reliable applications. Despiterecent empirical studies positing a near-perfect linear correlation betweenin-distribution (ID) and out-of-distribution (OOD) accuracies, we empiricallydemo