Learning Causal Structures Using Regression Invariance

Authors: AmirEmad Ghassami, Saber Salehkaleybar, Negar Kiyavash, Kun Zhang

NeurIPS 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiment results show that the proposed algorithm outperforms the other existing algorithms. We evaluate the performance of LRE algorithm by testing it on both synthetic and real data.
Researcher Affiliation Academia Department of ECE, University of Illinois at Urbana-Champaign, Urbana, USA. Coordinated Science Laboratory, University of Illinois at Urbana-Champaign, Urbana, USA. Department of Philosophy, Carnegie Mellon University, Pittsburgh, USA.
Pseudocode Yes Algorithm 1 The Baseline Algorithm and Algorithm 2 LRE Algorithm are provided.
Open Source Code No The paper mentions using the 'pcalg package [15]' but does not provide a link or statement about releasing the source code for their own proposed methodology (LRE algorithm) or baseline.
Open Datasets Yes We considered dataset of educational attainment of teenagers [27]. We considered GRNs in DREAM 3 In Silico Network challenge, conducted in 2008 [19]. The structures of these networks are available in the open-source tool Gene Net Weaver (GNW) [28].
Dataset Splits No The paper discusses generating samples for experiments and splitting real data into different 'environments', but it does not specify explicit training, validation, or test dataset splits with percentages, counts, or a detailed methodology for reproduction.
Hardware Specification No The paper does not provide specific hardware details (e.g., CPU, GPU models, or cloud instance types) used for running its experiments.
Software Dependencies No The paper mentions using 'pcalg package [15]' and 'Gene Net Weaver (GNW)' [28], but it does not provide specific version numbers for these or other software dependencies.
Experiment Setup Yes We generated data from a linear Gaussian SEM with coefficients drawn uniformly at random from [0.1, 2], and the variance of each exogenous noise was drawn uniformly at random from [0.1, 4]. In order to obtain measurements from the second environment, we increased coefficients of exogneous noise terms from 0.05 to 0.2 in GNW tool. We ran LRE algorithm on the two parts of data as two environments with a significance level of 0.01.