Graph metrics as summary statistics for Approximate Bayesian Computation with application to network model parameter estimation
Graph metrics as summary statistics for Approximate Bayesian Computation with application to network model parameter estimation
复制标题
DOI：
10.1093/comnet/cnu009
复制
发表时间：
2015-03-01
影响因子：
2.1
通讯作者：
Jurman, Giuseppe
中科院分区：
文献类型：
--
作者：
Fay, Damien;Moore, Andrew W.;Jurman, Giuseppe
在本文中，我们研究近似贝叶斯计算作为一种技术，估计相对于观察到的图的图生成器的参数。具体而言，我们调查了六个谱图指标，以评估其适用性作为汇总统计。总体结果是，近似贝叶斯计算可以导致合理的估计参数后验，如果度量的秩是足够高的。对于某些图形度量，估计参数中可能存在偏差，尽管这些偏差在经验上看起来很小。我们证明，结合指标，形成一个新的汇总统计提供了更强大的估计。鉴于这些结果，作者然后创建两个，有点任意，图形生成器，并显示如何这些参数可以轻松地估计。此外，我们展示了如何应用模型选择来确定哪个生成器最好地解释了观察到的图形。
In this paper, we investigate Approximate Bayes Computation as a technique for estimating the parameters of graph generators relative to an observed graph. Specifically, we investigate six spectral graph metrics with a view to evaluating their suitability as summary statistics. The overall findings are that Approximate Bayesian Computation can result in reasonable estimates of the parameter posteriors, if the rank of the metrics is sufficiently high. For some graph metrics, biases can exist in the estimated parameters though these appear, empirically, to be small. We demonstrate that combining metrics to form a new summary statistic provides more robust estimates. Given these results, the authors then create two, somewhat arbitrary, graph generators and show how the parameters for these may be estimated with ease. In addition, we show how to apply model selection to determine which generator best explains the observed graph.