arXiv：cs.LG· Shun Yanashima, Kentaro Kanamori, Hirofumi Suzuki·· 4 小时前AI 评分28
AdaPS-LiNGAM：小样本场景下线性非高斯无环模型的自适应前驱选择
AdaPS-LiNGAM: Adaptive Predecessor Selection for Linear Non-Gaussian Acyclic Models under Small-Sample Settings
AI 导读
AdaPS-LiNGAM 通过自适应选择稀疏前驱子集重建残差，解决了 DirectLiNGAM 在变量数超过样本量时残差化退化的结构性问题。该方法将同一子集选择原则用于边估计的最终剪枝步骤，在合成数据的小样本场景下能准确恢复因果结构，且随样本量减少性能下降更平缓。
正文
Abstract:Causal discovery becomes particularly challenging when the available sample size is small relative to the number of variables. This challenge also arises in the linear non-Gaussian acyclic model (LiNGAM), an identifiable framework for causal discovery from observational data. DirectLiNGAM estimates a causal order, which arranges variables so that causes precede their effects, by sequentially identifying an exogenous variable and removing its linear effect from the remaining variables. We establish a structural limitation of this procedure: when the number of variables exceeds the sample size, repeated residualization necessarily becomes degenerate before the full causal order can be determined. Our analysis further reveals that each residual can be reconstructed using only a graph-determined subset of variables already placed earlier in the causal order, termed the active boundary. This result motivates AdaPS-LiNGAM (Adaptive Predecessor Selection LiNGAM), which reconstructs each residual directly from the original observations using an adaptively chosen sparse subset of those earlier variables. The same subset-selection principle is also applied to the final pruning step for edge estimation. Experiments on synthetic data demonstrate that AdaPS-LiNGAM provides accurate causal-structure recovery in sample-limited settings and degrades more gradually as the sample size decreases.
| Subjects: | Machine Learning (cs.LG); Machine Learning (stat.ML) |
| Cite as: | arXiv:2610.09782 [cs.LG] |
| (or arXiv:2610.09782v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.09782 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Shun Yanashima [view email]
[v1]
Wed, 7 Oct 2026 10:01:25 UTC (266 KB)
来源：arXiv：cs.LG · arxiv.org