Fast Estimation of Causal Interactions using Wold Processes
Fast Estimation of Causal Interactions using Wold Processes
复制标题
使用 Wold 过程快速估计因果交互作用
DOI：
--
复制
发表时间：
2018
期刊：
影响因子：
--
通讯作者：
R. Assunção
中科院分区：
文献类型：
--
作者：
F. Figueiredo;Guilherme R. Borges;Pedro O. S. Vaz de Melo;R. Assunção
本文主要研究多元点过程的Granger因果矩阵的学习问题。为了完成这项任务，我们的工作是第一次探索使用Wold过程。通过这样做，我们能够开发出渐近快速的MCMC学习算法。在事件总数为$N$、进程数为$K$的情况下，我们的学习算法每次迭代的代价为$O(N(\，\log(N)，+\，\log(K)。这比目前的$O(N^3\，K^2)$或$O(K^3)$要快得多。我们的方法被称为GrangerBusca，在九个数据集上进行了验证。这与大多数以前的工作相比是一个进步，这些工作主要集中在Memetracker数据的子集上。在精确度方面，GrangerBusca的精确度是通常探索的Memetracker子集的三倍(精度@10)。由于GrangerBusca的训练复杂性要低得多，因此我们的方法是唯一能够为更大、完整的数据集训练模型的方法。
We here focus on the task of learning Granger causality matrices for multivariate point processes. In order to accomplish this task, our work is the first to explore the use of Wold processes. By doing so, we are able to develop asymptotically fast MCMC learning algorithms. With $N$ being the total number of events and $K$ the number of processes, our learning algorithm has a $O(N(\,\log(N)\,+\,\log(K)))$ cost per iteration. This is much faster than the $O(N^3\,K^2)$ or $O(K^3)$ for the state of the art. Our approach, called GrangerBusca, is validated on nine datasets. This is an advance in relation to most prior efforts which focus mostly on subsets of the Memetracker data. Regarding accuracy, GrangerBusca is three times more accurate (in Precision@10) than the state of the art for the commonly explored subsets Memetracker. Due to GrangerBusca's much lower training complexity, our approach is the only one able to train models for larger, full, sets of data.
影响因子：
8.6
作者：
Dhamala, Mukeshwar;Rangarajan, Govindan;Ding, Mingzhou
通讯作者：
Ding, Mingzhou