从不完整数据中同时学习有环因果图和缺失机制。
MissNODAG: Differentiable Cyclic Causal Graph Learning from Incomplete Data
- 用可微分框架联合优化因果结构与缺失机制
- 在真实基因扰动数据上准确恢复有环因果关系
- 适用于缺失非随机的复杂系统分析
现实世界系统（如生物网络）的因果发现常受反馈环和不完整数据干扰。传统算法假设无环结构或数据完全可观测，难以应对此类挑战。为此，我们提出 MissNODAG，一个可微分框架，能够从部分观测数据（包括缺失非随机情况）中同时学习底层有环因果图和缺失机制。该框架结合加性噪声模型与期望最大化过程，在缺失值插补与观测数据似然优化之间交替迭代，以揭示因果结构与缺失模式。我们在大样本下建立了得分函数精确最大化时的一致性保证。通过合成实验与真实基因扰动数据应用，验证了 MissNODAG 的有效性。
原文摘要 · Abstract (English)
Causal discovery in real-world systems, such as biological networks, is often complicated by feedback loops and incomplete data. Standard algorithms, which assume acyclic structures or fully observed data, struggle with these challenges. To address this gap, we propose MissNODAG, a differentiable framework for learning both the underlying cyclic causal graph and the missingness mechanism from partially observed data, including data missing not at random. Our framework integrates an additive noise model with an expectation-maximization procedure, alternating between imputing missing values and optimizing the observed data likelihood, to uncover both the cyclic structures and the missingness mechanism. We establish consistency guarantees under exact maximization of the score function in the large sample setting. Finally, we demonstrate the effectiveness of MissNODAG through synthetic experiments and an application to real-world gene perturbation data.
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。