Hierarchical network models for exchangeable structured interaction processes.
Hierarchical network models for exchangeable structured interaction processes.
复制标题
用于可交换结构化交互过程的分层网络模型。
DOI：
10.1080/01621459.2021.1896526
复制
发表时间：
2022
影响因子：
3.7
通讯作者：
Hero,Alfred
中科院分区：
文献类型：
--
作者：
Dempsey,Walter;Oselio,Brandon;Hero,Alfred
网络数据通常是通过一系列结构化的相互作用而产生的。例如，电子邮件交换有一个发送者，后面可能有多个接收者。另一方面，科学文章可能有多个学科领域和多个作者。我们引入了一个统计模型，称为Pitman-Yor层次顶点组件模型（PY-HVCM），这是非常适合结构化的交互数据。所提出的PY-HVCM有效地模拟复杂的关系数据的局部汇集的信息，通过一个潜在的，共享的人口水平的分布。PY-HCVM是层次顶点组件模型的一个典型例子，它是可交换结构化交互标记网络模型的一个子类，即网络对交互重新标记不变性。理论分析和支持仿真提供了清晰的模型解释，并建立了全局稀疏度和幂律度分布。推导了一个计算上易于处理的Gibbs抽样算法，用于推断复杂网络的稀疏性和幂律性质。我们在安然电子邮件数据集和ArXiv数据集上演示了该模型，通过后验预测验证显示了该模型的拟合优度。
Network data often arises via a series ofstructured interactionsamong a population of constituent elements. E-mail exchanges, for example, have a single sender followed by potentially multiple receivers. Scientific articles, on the other hand, may have multiple subject areas and multiple authors. We introduce a statistical model, termed the Pitman-Yor hierarchical vertex components model (PY-HVCM), that is well suited for structured interaction data. The proposed PY-HVCM effectively models complex relational data by partial pooling of local information via a latent, shared population-level distribution. The PY-HCVM is a canonical example ofhierarchical vertex components models—a subfamily of models forexchangeable structured interaction-labeled networks, that is, networks invariant to interaction relabeling. Theoretical analysis and supporting simulations provide clear model interpretation, and establish global sparsity and power law degree distribution. A computationally tractable Gibbs sampling algorithm is derived for inferring sparsity and power law properties of complex networks. We demonstrate the model on both the Enron e-mail dataset and an ArXiv dataset, showing goodness of fit of the model via posterior predictive validation.