在存在隐变量混杂时，用观测数据和因果图就能准确做因果推断。
DeCaFlow: A deconfounding causal generative model
- 基于观测数据和因果图，用代理变量修正混杂偏差
- 在3个隐变量混杂的Ecoli70数据集上，对数百个因果问题准确估计
- 无需重新训练，直接适配任意给定的因果图，适合实际应用
我们提出DeCaFlow，一种去混杂的因果生成模型。仅需一次针对每个数据集的训练，利用观测数据和底层因果图，即可在存在隐性混杂因素的情况下对连续变量进行准确的因果推断。具体而言，我们扩展了已有成果，证明单个DeCaFlow实例能正确估计所有可通过do-演算识别的因果查询，当do-演算不足时，借助代理变量调整因果效应。此外，我们证明反事实查询只要其干预版本可识别，就同样可被正确估计。在多种场景下的实证结果表明，包括含有3个独立隐性混杂因子、数十个可观测变量和数百个因果查询的Ecoli70数据集，DeCaFlow优于现有方法，并展现出即插即用的适用性，支持任意给定的因果图。代码见https://github.com/aalmodovares/DeCaFlow。
原文摘要 · Abstract (English)
We introduce DeCaFlow, a deconfounding causal generative model. Training once per dataset using just observational data and the underlying causal graph, DeCaFlow enables accurate causal inference on continuous variables under the presence of hidden confounders. Specifically, we extend previous results on causal estimation under hidden confounding to show that a single instance of DeCaFlow provides correct estimates for all causal queries identifiable with do-calculus, leveraging proxy variables to adjust for the causal effects when do-calculus alone is insufficient. Moreover, we show that counterfactual queries are identifiable as long as their interventional counterparts are identifiable, and thus are also correctly estimated by DeCaFlow. Our empirical results on diverse settings (including the Ecoli70 dataset, with 3 independent hidden confounders, tens of observed variables and hundreds of causal queries) show that DeCaFlow outperforms existing approaches, while demonstrating its out-of-the-box applicability to any given causal graph. An implementation can be found in https://github.com/aalmodovares/DeCaFlow
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。