Efficiently Finding Conditional Instruments for Causal Inference

Authors: Benito van der Zander, Johannes Textor, Maciej Liskiewicz

IJCAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical Here we address two issues with conditional IVs. First, they are conceptually rather distant to standard IVs; even variables that are independent of X could qualify as conditional IVs. We propose a new concept called ancestral IV, which interpolates between the two existing notions. Second, so far only exponential-time algorithms are known to find conditional IVs in a given causal diagram. Indeed, we prove that this problem is NP-hard. Nevertheless, we show that whenever a conditional IV exists, so does an ancestral IV, and ancestral IVs can be found in polynomial time.
Researcher Affiliation Academia Benito van der Zander University of L ubeck Germany benito@tcs.uni-luebeck.de Johannes Textor Utrecht University The Netherlands johannes.textor@gmx.de Maciej Li skiewicz University of L ubeck Germany liskiewi@tcs.uni-luebeck.de
Pseudocode Yes function NEAREST-SEPARATOR(G, Y, Z) [...] function ANCESTRAL-INSTRUMENT(G, X, Y, Z)
Open Source Code No Our algorithms are implemented in the open-source software DAGitty [Textor et al., 2011]. (This refers to existing software that implements their algorithms, not a release of new source code specifically for the paper's methodology.)
Open Datasets No The paper is theoretical and does not involve training models on datasets. Therefore, it does not provide access information for a dataset used for training.
Dataset Splits No The paper is theoretical and does not involve empirical validation on datasets. Therefore, it does not discuss training/validation/test dataset splits.
Hardware Specification No The paper is theoretical and focuses on algorithms and proofs rather than empirical experiments, so it does not specify any hardware used.
Software Dependencies No The paper mentions "DAGitty [Textor et al., 2011]" as open-source software where their algorithms are implemented, but it does not specify version numbers for DAGitty or any other software dependencies required to replicate the theoretical work or algorithms presented.
Experiment Setup No The paper is theoretical and focuses on algorithms and proofs rather than empirical experiments. Thus, it does not provide details about an experimental setup, hyperparameters, or training settings.