arXiv:2502.06547stat.MLcs.LG2025-02被引 3
通过数据增强与正则化，让模型自动学习对称性不变性。
Data Augmentation and Regularization for Learning Group Equivariance
- 结合数据增强与正则化项，引导模型学习群等变性。
- 在多种对称变换下，模型预测结果保持一致。
- 适合需要提升泛化能力的对称结构建模任务。
在许多机器学习任务中，已知的对称性可作为归纳偏置以提升模型性能。本文总结了我们先前工作中的成果，并将其拓展至证明：通过在增强数据上训练并辅以正则化，可实现模型的群等变性。该方法在多种对称变换下均能有效提升模型鲁棒性，且无需显式编码对称结构。
原文摘要 · Abstract (English)
In many machine learning tasks, known symmetries can be used as an inductive bias to improve model performance. In this paper, we consider learning group equivariance through training with data augmentation. We summarize results from a previous paper of our own, and extend the results to show that equivariance of the trained model can be achieved through training on augmented data in tandem with regularization.
数据增强等变性正则化
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。