研究表明不均衡训练数据也能实现稳健的子群体泛化
Guiding Data Allocation for Robust Subpopulation Generalization
这篇论文从理论上推翻了“训练数据必须各子群体完全平衡”的默认假设，证明明显不均衡的数据配置也能达到相当的准确率，还给出了按梯度方向调整比例的实用方法。做数据采购标注预算紧张的机器学习工程师、尤其是处理人脸或医疗等存在子群体偏移数据的团队，可以据此省下不少数据收集成本。
一篇新论文研究了子群体偏移下平衡训练数据是否为唯一最优配置。作者扩展高斯特征模型框架，从理论上刻画了不同训练子群体比例下的性能分布，并证明包括明显不均衡配置在内的多种方案，可以取得与完全平衡相当的表现。
论文进一步推导出基于梯度的方向，用于调整子群体比例，以更高效地提升调整后准确率与最差组准确率。作者据此认为，策略性地分配数据比简单平衡更灵活、更有原则。代码已开源在 GitHub 仓库 zhaoyingpan/subpop-data-allocation。对受采集成本限制、难以做到各子群体完全均衡的团队而言，这一结论提供了降低数据收集成本的可行路径。
Guiding Data Allocation for Robust Subpopulation Generalization
Machine learning models often achieve high average accuracy in the training data while performing poorly on test data due to subpopulation shift. A common data-centric solution is to enforce balanced subgroup proportions, but full balance can be costly, infeasible, and not always necessary. In this work, we study whether balanced training data is the unique optimal configuration for robust subpopulation generalization. We extend a Gaussian feature-model framework to theoretically characterize the performance landscape across general training subgroup distributions. We show that multiple configurations, including substantially imbalanced ones, can perform comparably to the balanced one. We further derive gradient-based directions for adjusting subgroup proportions to improve adjusted accuracy and worst-group accuracy more efficiently. These findings suggest that strategically allocating data offers a more flexible and principled path to robust performance than simple balancing. The code is available at https://github.com/zhaoyingpan/subpop-data-allocation.
This report is categorized under AITOP's "Papers" desk. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include 子群体偏移, 数据分配, 机器学习, 鲁棒性.
Source attribution: curated from arXiv cs.LG (@Zhaoying Pan, Yipei Wang, Shenyu Lu, Xiaoqian Wang).