Showing 1–1 of 1 results for author: Gössler, G

Identifiability of total effects from abstractions of time series causal graphs

Authors: Charles K. Assaad, Emilie Devijver, Eric Gaussier, Gregor Gössler, Anouar Meynaoui

Abstract: We study the problem of identifiability of the total effect of an intervention from observational time series in the situation, common in practice, where one only has access to abstractions of the true causal graph. We consider here two abstractions: the extended summary causal graph, which conflates all lagged causal relations but distinguishes between lagged and instantaneous relations, and the… ▽ More We study the problem of identifiability of the total effect of an intervention from observational time series in the situation, common in practice, where one only has access to abstractions of the true causal graph. We consider here two abstractions: the extended summary causal graph, which conflates all lagged causal relations but distinguishes between lagged and instantaneous relations, and the summary causal graph which does not give any indication about the lag between causal relations. We show that the total effect is always identifiable in extended summary causal graphs and provide sufficient conditions for identifiability in summary causal graphs. We furthermore provide adjustment sets allowing to estimate the total effect whenever it is identifiable. △ Less

Submitted 16 February, 2025; v1 submitted 23 October, 2023; originally announced October 2023.

Comments: Correction to original version published at UAI 2024. The published version contains an error in Condition 2(c) in Theorem 2, which is corrected in this version