A polynomial-time algorithm for learning nonparametric causal graphs

Authors: Ming Gao, Yi Ding, Bryon Aragam

NeurIPS 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Finally, we compare the proposed algorithm to existing approaches in a simulation study. Our simulation results can be summarized as follows: When implemented using generalized additive models [19], our method outperforms most state-of-the-art methods, particularly on denser graphs with hub nodes.
Researcher Affiliation Academia Ming Gao University of Chicago minggao@uchicago.edu Yi Ding University of Chicago dingy@uchicago.edu Bryon Aragam University of Chicago bryon@chicagobooth.edu
Pseudocode Yes Algorithm 1 Population algorithm for learning nonparametric DAGs. Algorithm 2 NPVAR algorithm.
Open Source Code Yes Code implementing the NPVAR algorithm is publicly available at https://github.com/Ming Gao97/ NPVAR.
Open Datasets No The paper describes generating synthetic data based on specified Graph types (Markov chains, Erdös-Rényi graphs, scale-free graphs) and Probability models (additive GPs, non-additive GPs, sine model, GLM). It does not use pre-existing publicly available datasets or provide links to them. Therefore, there is no concrete access information for a publicly available dataset.
Dataset Splits No The paper does not explicitly provide training/validation/test dataset splits. It mentions "Randomly split the n samples in half" in Algorithm 2, which is an internal sample splitting technique for estimation, not a formal dataset partitioning for reproducibility of overall experimental results.
Hardware Specification No The paper does not provide any specific details about the hardware used for running the experiments, such as GPU or CPU models.
Software Dependencies No The paper mentions using "generalized additive models (GAMs)" for implementation but does not specify any software names with version numbers (e.g., Python, PyTorch, specific GAMs library with version).
Experiment Setup No The paper describes the graph types and probability models used for simulation studies but does not provide specific hyperparameter values or detailed system-level training settings in the main text. It states: "Full details of the implementations used as well as additional experiments can be found in the supplement."