Critique of: “A Parallel Framework for Constraint-Based Bayesian Network Learning via Markov Blanket Discovery” by SCC Team From UC San Diego
Critique of: “A Parallel Framework for Constraint-Based Bayesian Network Learning via Markov Blanket Discovery” by SCC Team From UC San Diego
复制标题
加州大学圣地亚哥分校 SCC 团队对“通过马尔可夫毯子发现进行基于约束的贝叶斯网络学习的并行框架”的评论
DOI：
10.1109/tpds.2022.3217284
复制
发表时间：
2023
影响因子：
5.3
通讯作者：
Rodriguez, Paul
中科院分区：
文献类型：
--
作者：
Gupta, Arunav;Ge, John;Li, John;Kong, Zihao;He, Kaiwen;Mikhailov, Matthew;Chin, Bryan;Li, Xiaochen;Apodaca, Max;Rodriguez, Paul
近年来，贝叶斯网络（BN）已成为流行的描述自然现象的情况下，因果关系是很重要的理解。为了解决学习贝叶斯网络固有的不可追踪性，Srivastava等人在他们题为“A Parallel Framework for Constraint-based Bayesian Network Learning via Markov Blanket Discovery”的论文中提出了一种基于马尔可夫毯发现的学习方法。我们能够再现强和弱缩放实验从纸上到128个核心，并验证通信成本缩放的所有三种算法的文件。我们还介绍了方法上的改进，弱标度，表明该论文的研究结果是独特的方法，而不是使用的数据集。由于数据集、核心计数和作业调度的差异，观察到性能的轻微变化。
Bayesian networks (BNs) have become popular in recent years to describe natural phenomena in situations where causal linkages are important to understand. In order to get around the inherent non-tractability of learning BNs, Srivastava et al. propose a markov blanket discovery-based approach to learning in their paper titled“A Parallel Framework for Constraint-based Bayesian Network Learning via Markov Blanket Discovery.”We are able to reproduce both the strong and weak scaling experiments from the paper up to 128 cores, and verify communication cost scaling for all three algorithms in the paper. We also introduce methodological improvements to weak scaling that show the paper's findings are unique to the methodology and not the datasets used. Slight variations in performance were observed due to differences in datasets, core count, and job scheduling.