arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2510.04263cs.LGcs.AI
收藏

高效潜变量因果发现:结合评分搜索与定向检验

Efficient Latent Variable Causal Discovery: Combining Score Search and Targeted Testing

Department of Philosophy(哲学系);Carnegie Mellon University(卡内基梅隆大学);Department of Psychiatry & Behavioral Sciences(精神病学与行为科学系);University of Minnesota(明尼苏达大学)

Joseph Ramsey, Bryan Andrews, Peter Spirtes

AI总结:

本文提出评分引导的混合因果搜索算法族,包括 BOSS-FCI、GRaSP-FCI、FCIT 和轻量级 LV-Dumb,通过用定向评分检验替代穷举检验,显著提升潜变量因果发现的效率与正确性。

摘要:

当存在潜变量或选择偏差时,从观测数据中学习因果结构尤其具有挑战性。快速因果推断(FCI)算法针对这一场景,但在许多子集上执行穷举式条件独立性检验,往往导致虚假独立性、缺失或多余边以及不可靠的定向。我们提出一族评分引导的混合策略因果搜索算法来扩展该框架。首先,我们引入 BOSS-FCI 和 GRaSP-FCI,它们是 GFCI(贪婪快速因果推断)的变体,分别用 BOSS(最佳顺序评分搜索)或 GRaSP(最稀疏排列的贪婪松弛)替代 FGES(快速贪婪等价搜索),在保持正确性的同时权衡可扩展性与保守性。其次,我们开发了 FCI 定向检验(FCIT),这是一种新颖的混合方法,用由 BOSS 引导的定向、评分信息检验取代穷举检验。FCIT 保证生成良构的部分祖先图(PAG),并在不同样本量下实现更高的精度和效率。最后,我们提出一种轻量级启发式方法 LV-Dumb(潜变量“哑”方法),它返回 BOSS 有向无环图(DAG)的 PAG。尽管在潜混杂情形下并非严格可靠,LV-Dumb 通常能达到与 FCIT 相当的准确率,同时运行速度显著更快。模拟和真实数据分析表明,BOSS-FCI 和 GRaSP-FCI 提供了稳健的基线,FCIT 在精度与可靠性之间取得最佳平衡,而 LV-Dumb 提供了一种快速、近乎等效的替代方案。综合来看,这些方法表明定向且评分引导的策略能够显著提升潜变量因果发现的效率和正确性。

英文摘要:

Learning causal structure from observational data is especially challenging when latent variables or selection bias are present. The Fast Causal Inference (FCI) algorithm addresses this setting but performs exhaustive conditional independence tests across many subsets, often leading to spurious independences, missing or extra edges, and unreliable orientations. We present a family of score-guided mixed-strategy causal search algorithms that extend this framework. First, we introduce BOSS-FCI and GRaSP-FCI, variants of GFCI (Greedy Fast Causal Inference) that substitute BOSS (Best Order Score Search) or GRaSP (Greedy Relaxations of Sparsest Permutation) for FGES (Fast Greedy Equivalence Search), preserving correctness while trading off scalability and conservativeness. Second, we develop FCI Targeted-Testing (FCIT), a novel hybrid method that replaces exhaustive testing with targeted, score-informed tests guided by BOSS. FCIT guarantees well-formed PAGs and achieves higher precision and efficiency across sample sizes. Finally, we propose a lightweight heuristic, LV-Dumb (Latent Variable "Dumb"), which returns the PAG of the BOSS DAG (Directed Acyclic Graph). Though not strictly sound for latent confounding, LV-Dumb often matches FCIT's accuracy while running substantially faster. Simulations and real-data analyses show that BOSS-FCI and GRaSP-FCI provide robust baselines, FCIT yields the best balance of precision and reliability, and LV-Dumb offers a fast, near-equivalent alternative. Together, these methods demonstrate that targeted and score-guided strategies can dramatically improve the efficiency and correctness of latent-variable causal discovery.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑