提出工具包与理论框架，解决数据来源变化导致的模型失效问题。
DeconDTN-Toolkit: A Library for Evaluation and Enhancement of Robustness to Provenance Shift
- 建立溯源漂移与不变学习的理论联系，设计鲁棒性学习目标。
- 发现传统方法在溯源漂移下性能显著下降，提出新评估指标。
- 提供可复用工具包，适配现有基准，支持算法验证与修复。
尽管分布漂移研究日益丰富，数据源与标签关系在部署时发生变化的溯源漂移问题仍缺乏理解与应对。本文首次建立溯源漂移、反事实不变性与不变学习之间的形式化关联，推导出提升鲁棒性的学习目标。随后提出DeconDTN-Toolkit，一个专门用于模拟不同强度溯源漂移的评估与修复工具套件，保持原有训练协议和基准基础设施不变。实验揭示经验风险最小化在溯源漂移下的脆弱性，引入一种新的泛化性能评估指标，并对现有算法进行了全面评估。本工作提供了刻画溯源混淆问题的理论基础与实用工具，以及相应的缓解方法实现。
原文摘要 · Abstract (English)
Despite the burgeoning body of work on distribution shifts, provenance shift-where the relationship between data source and label changes at deployment-remains poorly understood and under-addressed. In this paper, we establish a formal connection between provenance shift, counterfactual invariance, and invariant learning to derive a learning objective for robustness. We then introduce \textsc{DeconDTN-Toolkit}, a specialized evaluation and remediation suite designed to simulate provenance shifts of varying degrees while maintaining the training protocol and the infrastructure of existing benchmarks. We reveal the vulnerability of Empirical Risk Minimization under provenance shift, introduce a robust out-of-distribution performance indicator, and conduct a comprehensive evaluation on existing algorithms. Our work provides both the theoretical grounding and the practical tools necessary to characterize the problem of confounding by provenance, and implementations of methods to mitigate it.
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。