Parallel Bayesian Network Structure Learning

Authors: Tian Gao, Dennis Wei

ICML 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We test the proposed PSL algorithm with different numbers K of agents and compare to the baseline GGSL (Gao et al., 2017) algorithm on the benchmark ALARM dataset.
Researcher Affiliation Industry Tian Gao* 1 Dennis Wei* 1 1IBM Research, Yorktown Heights, NY 10598 USA.
Pseudocode Yes Algorithm 1 Local Learn, Algorithm 2 Parallel BN Structure Learning, Algorithm 3 Parallel Local Learn
Open Source Code No The paper does not provide an explicit statement or link for the open-source code of the proposed methodology.
Open Datasets Yes We test the proposed PSL algorithm... on the benchmark ALARM dataset.
Dataset Splits No We use 1000 samples from the dataset and perform structure learning with different algorithms.
Hardware Specification Yes The experiments are conducted on a machine with a 2.3GHz Intel i5-5300U processor.
Software Dependencies No We use an existing Dynamic Programming algorithm (Silander & Myllymaki, 2006) as the BNStruct Learn subroutine for all experiments.
Experiment Setup No We use 1000 samples from the dataset and perform structure learning with different algorithms. ... We fix p = 0.7 for target selection function choose Target.