arXiv cs.CL· Yaroslav Kivva, Aliaksandra Novik, Negar Kiyavash·· 11 小时前AI 评分24
研究揭示 do-calculus 在 Universal Identifiability 问题上的不完备性
On the Incompleteness of do-calculus for Universal Identifiability
AI 导读
研究提出 do-calculus 完备性的严格形式化定义，并证明此前 ID 与 gID 问题中的完备性结论依然成立。但在更一般的 Universal Identifiability（uID）问题上，do-calculus 的规则不再完备，这可能是该问题至今悬而未决的原因。
正文 · 原文
Abstract:We study the problem of causal effect identification from an arbitrary set of observational and/or interventional distributions, which we call Universal Identifiability (uID). The uID problem generalizes existing classical identifiability (ID) and general identifiability (gID) problems, which considered the identification problem from either mere observational distributions or a specific type of interventional distributions. Prior research has established graphical criteria to determine when a causal effect is identifiable in ID and gID problems. Moreover, algorithms that leverage do-calculus and basic probability manipulations have been proposed to compute these effects. Consequently, it has been shown that do-calculus is both sound and complete in these problems. Despite these well-established results, a formal definition of what it means for the do-calculus to be complete does not exist. We propose a rigorous formalization for the completeness of the do-calculus and show that the completeness results in previous work remain valid. However, for more general problems, such as uID, the rules of do-calculus are no longer complete. This is perhaps the reason why the uID problem has remained open thus far.
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Probability (math.PR) |
| Cite as: | arXiv:2611.08916 [cs.CL] |
| (or arXiv:2611.08916v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2611.08916 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Yaroslav Kivva [view email]
[v1]
Fri, 6 Nov 2026 02:59:26 UTC (38 KB)
来源：arXiv cs.CL · arxiv.org