arXiv cs.AI· Zeqin Yang, Weilin Chen, Ruichu Cai, Yuguang Yan, Zhifeng Hao, Zhipeng Yu, Zhichao Zou, Jixing Xu, Zhen Peng, Jiecheng Guo·· 4 小时前AI 评分29
利用最优传输权重估计长期异质剂量-反应曲线的泛化边界
Estimating Long-term Heterogeneous Dose-response Curve: Generalization Bound Leveraging Optimal Transport Weights
AI 导读
一篇 arXiv 论文研究在存在未观测混杂因子和连续处理的条件下，估计长期异质剂量-反应曲线（HDRC）。方法上，论文引入最优传输加权框架，将长期观测数据对齐到辅助的短期实验数据以去除未观测混杂因子，并基于重加权分布建立反事实预测误差的泛化边界，进而构造长期 HDRC 估计器。合成与半合成数据集上的大量实验验证了方法的有效性。
正文 · 原文
Abstract:Long-term treatment effect estimation is a significant but challenging problem in many applications. Existing methods rely on ideal assumptions, such as no unobserved confounders or binary treatment, to estimate long-term average treatment effects. However, in numerous real-world applications, these assumptions could be violated, and average treatment effects are insufficient for personalized decision-making. In this paper, we address a more general problem of estimating long-term Heterogeneous Dose-Response Curve (HDRC) while accounting for unobserved confounders and continuous treatment. Specifically, to remove the unobserved confounders in the long-term observational data, we introduce an optimal transport weighting framework to align the long-term observational data to an auxiliary short-term experimental data. Furthermore, to accurately predict the heterogeneous effects of continuous treatment, we establish a generalization bound on counterfactual prediction error by leveraging the reweighted distribution induced by optimal transport. Finally, we develop a long-term HDRC estimator building upon the above theoretical foundations. Extensive experiments on synthetic and semi-synthetic datasets demonstrate the effectiveness of our approach.
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2406.19195 [cs.LG] |
| (or arXiv:2406.19195v4 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2406.19195 arXiv-issued DOI via DataCite |
Submission history
From: Zeqin Yang [view email]
[v1]
Thu, 27 Jun 2024 14:13:46 UTC (1,304 KB)
[v2]
Fri, 16 May 2025 07:28:31 UTC (1,319 KB)
[v3]
Wed, 22 Oct 2025 13:52:31 UTC (1,131 KB)
[v4]
Mon, 26 Oct 2026 09:12:07 UTC (1,168 KB)
来源：arXiv cs.AI · arxiv.org