Entropic Causal Inference: Graph Identifiability

Authors: Spencer Compton, Kristjan Greenewald, Dmitriy A Katz, Murat Kocaoglu

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

Reproducibility Variable Result LLM Response
Research Type Experimental We rigorously evaluate the performance of our algorithms on synthetic data generated from a variety of models, observing improvement over prior work. Finally we test our algorithms on real-world datasets.
Researcher Affiliation Collaboration 1Massachusetts Institute of Technology, Cambridge, USA 2MIT-IBM Watson AI Lab, Cambridge, USA 3IBM Research, Cambridge, USA 4Purdue University, West Lafayette, USA.
Pseudocode Yes Algorithm 1 Learning general graphs with oracle
Open Source Code No The paper refers to an external repository (bnlearn) which contains data, but does not provide a link or statement about open-sourcing the code for the methodology described in this paper.
Open Datasets Yes We also apply our algorithms on semi-synthetic data using the bnlearn1 repository and demonstrate the applicability of low-entropy assumptions and the proposed method. 1https://www.bnlearn.com/bnrepository/
Dataset Splits No The paper does not provide explicit training/test/validation dataset splits (e.g., percentages or sample counts). It mentions using synthetic and real-world data but not how they are partitioned for different stages.
Hardware Specification No The paper does not explicitly describe the hardware used to run its experiments (e.g., specific GPU/CPU models, memory details).
Software Dependencies No The paper mentions 'bnlearn' and 'R' but does not provide specific version numbers for these or other software dependencies.
Experiment Setup Yes We evaluate performance via the structural Hamming distance (SHD) from the estimated graph to the true causal graph. See the Appendix for implementation details. The x axis shows entropy of the exogenous noise. The exogenous noise of the first variable is fixed to be large ( 3.3 bits), hence it is a high entropy source (HES).