Neuropathic Pain Diagnosis Simulator for Causal Discovery Algorithm Evaluation

Authors: Ruibo Tu, Kun Zhang, Bo Bertilson, Hedvig Kjellstrom, Cheng Zhang

NeurIPS 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Using our simulator, we have evaluated extensively causal discovery algorithms under various settings.
Researcher Affiliation Collaboration Ruibo Tu KTH Royal Institute of Technology; Kun Zhang Carnegie Mellon University; Bo Christer Bertilson Karolinska Institute; Hedvig Kjellström KTH Royal Institute of Technology; Cheng Zhang Microsoft Research, Cambridge
Pseudocode No The paper describes the data generation process and parameter estimation, but it does not provide any pseudocode or clearly labeled algorithm blocks.
Open Source Code Yes The simulator is available at https://github.com/TURuibo/Neuropathic-Pain-Diagnosis-Simulator.
Open Datasets Yes To make our generated records close to the real-world scenario, we learn parameters from a dataset including 141 patient diagnostic records [46]. [46] C. Zhang, H. Kjellstrom, C. H. Ek, and B. C. Bertilson. Diagnostic prediction using discomfort drawings with IBTM. In MLHC, 2016.
Dataset Splits No The paper describes using a real-world dataset to estimate parameters and generating simulated data of various sample sizes for evaluation. However, it does not specify train, validation, and test splits for its own experimental setup or for the causal discovery algorithms evaluated.
Hardware Specification No The paper does not provide any specific details about the hardware used to run the experiments.
Software Dependencies No The paper mentions using "the causal discovery algorithms implemented by Tetrad [41]", but does not provide specific version numbers for Tetrad or any other software dependencies.
Experiment Setup No The paper describes how different practical issues (confounding, selection bias, missing data) are simulated and how causal discovery algorithms are applied under these conditions and varying sample sizes. However, it does not provide specific hyperparameters or system-level training settings for the causal discovery algorithms or any other models used in the experiments.