Identifiability by common backdoor in summary causal graphs of time series

2025-06-17

AI Summary

This paper addresses the identifiability of total causal effects in time-series settings with multiple interventions and multiple outcomes, under two realistic constraints: only an abstracted summary causal graph (not the full dynamic causal graph) is known, and only observational data are available. The goal is to determine whether the total effect can be estimated from observational data via a do-free formula—e.g., via backdoor adjustment. We establish necessary and sufficient conditions for *common-backdoor identifiability* for both time-varying and stationary time series, and propose a polynomial-time automated decision algorithm. Methodologically, we generalize the classical backdoor criterion and do-calculus to the abstract time-series modeling framework, thereby relaxing the requirement for full temporal graph structure. Our approach enables reliable, computationally tractable causal effect estimation even when structural knowledge is incomplete or coarse-grained.

Technology Category

• Reasoning under Uncertainty: Causality
• Machine Learning: Causal Learning
• Knowledge Representation and Reasoning: Action, Change, and Causality

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Web Mining and Content Analysis: Models for Web evolution
• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics

Abstract

The identifiability problem for interventions aims at assessing whether the total effect of some given interventions can be written with a do-free formula, and thus be computed from observational data only. We study this problem, considering multiple interventions and multiple effects, in the context of time series when only abstractions of the true causal graph in the form of summary causal graphs are available. We focus in this study on identifiability by a common backdoor set, and establish, for time series with and without consistency throughout time, conditions under which such a set exists. We also provide algorithms of limited complexity to decide whether the problem is identifiable or not.

Problem

• Assess total effect identifiability from observational data
• Study multiple interventions in time series graphs
• Establish conditions for common backdoor set existence

Innovation

• Identifiability via common backdoor set
• Conditions for time series consistency
• Algorithms for identifiability decision