A Core-Guided Approach to Learning Optimal Causal Graphs

Authors: Antti Hyttinen, Paul Saikko, Matti Järvisalo

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

Reproducibility Variable Result LLM Response
Research Type Experimental We compare the performance of Dseptor to those of state-of-art Max SAT solvers using the encoding of Hyttinen et al. [2014]. The experiments were run on 2.83-GHz Intel Xeon E5440 machines with 32-GB RAM and Debian GNU/Linux. First we compare the performance of the solvers on synthetic data generated from 7-node cyclic (possibly) linear Gaussian models with correlated disturbances. ... Figure 4 gives a comparison of Dseptor with the Max SAT solvers ... Dseptor outperforms all of the other solvers, solving 96% of the instances under the timeout of 600 seconds.
Researcher Affiliation Academia Antti Hyttinen and Paul Saikko and Matti J arvisalo HIIT, Department of Computer Science, University of Helsinki, Finland
Pseudocode Yes Alg. 1 outlines this approach as FINDCOREINCREMENTAL. ... Algorithm 1: Incremental core extraction in Dseptor.
Open Source Code No An implementation and an online appendix are available at the authors homepages." - This statement is too general and does not provide a specific, direct link to the source code for the methodology described in the paper, nor does it explicitly state the code is publicly released in a persistent repository.
Open Datasets Yes To examine the performances on more realistic sets of (in)dependence constraints, we looked at real-world datasets often used for benchmarking exact Bayesian network structure learning algorithms [Yuan and Malone, 2013; Bartlett and Cussens, 2017]." and Table 2 lists datasets such as "Adult", "Alarm", "Autos", "Bands", etc.
Dataset Splits No The paper mentions using N=1000 samples to obtain weights for independence constraints and refers to real-world datasets, but it does not specify explicit training, validation, or test dataset splits (e.g., percentages or counts) for reproducing its experiments.
Hardware Specification Yes The experiments were run on 2.83-GHz Intel Xeon E5440 machines with 32-GB RAM and Debian GNU/Linux.
Software Dependencies No We use Mini SAT [E en and S orensson, 2004] and IBM CPLEX as the internal SAT and IP solvers, respectively." - Specific version numbers for Mini SAT and IBM CPLEX are not provided.
Experiment Setup Yes We used BIC-based model selection to obtain the weights for the independence constraints over N = 1000 samples." and "We employed the BDEU score with equivalent sample size 10 to obtain constraint weights for this discrete data, and used a per-instance timeout of 7200 seconds." and "The edges were sampled at random to obtain average degree 2, coefficients were from [0.2, 0.8].