Partial Counterfactual Identification from Observational and Experimental Data

Authors: Junzhe Zhang, Jin Tian, Elias Bareinboim

ICML 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Our algorithms are validated extensively on synthetic and real-world datasets.
Researcher Affiliation Academia 1Department of Computer Science, Columbia University 2Department of Computer Science, Iowa State University.
Pseudocode Yes Algorithm 1 CREDIBLEINTERVAL
Open Source Code No The paper does not provide an explicit statement about the release of source code for the described methodology, nor does it include a link to a code repository.
Open Datasets Yes International stroke trials was a large, randomized, open trial of up to 14 days of antithrombotic therapy after stroke onset (Carolei et al., 1997). ... We collect N = 10^3 synthetic observational samples v = {x(n), y(n)}N n=1 that are compatible with the Bow diagram of Fig. 1d.
Dataset Splits No The paper mentions collecting 'N = 10^4 observational samples' or 'N = 10^3 observational samples' for different experiments, but it does not specify any training, validation, or test dataset splits.
Hardware Specification No The paper does not provide any specific details about the hardware (e.g., GPU/CPU models, memory) used to run the experiments.
Software Dependencies No The paper does not specify any software dependencies, libraries, or solvers with version numbers that would be needed to replicate the experiments.
Experiment Setup Yes In all experiments, we evaluate our proposed strategy using credible intervals (ci). We draw at least 4 x 10^3 samples from the posterior distribution P (θctf | v) over the target counterfactual. This allows us to compute 100% credible interval over θctf within error ϵ = 0.05, with probability at least 1 δ = 0.95.