论文标题
基于梯度的因果结构学习，并具有归一流的流量
Gradient-based Causal Structure Learning with Normalizing Flow
论文作者
论文摘要
在本文中，我们提出了一种基于得分的归一化流量方法，称为DAG-NF来学习输入观测数据的依赖性。受到计算机视觉中的Grad-CAM的启发，我们将输出的Jacobian矩阵作为因果关系，并且可以将此方法推广到任何神经网络，尤其是针对基于流动的生成神经网络（例如掩盖自动回应流（MAF））和持续的归一化流量（CNF），以计算对数目的可能性丢失和分配分配的分布数据和目标分配的分布。此方法扩展了符号，该符号对图节点的连续邻接矩阵强制执行重要的辅助约束，并显着降低了图形搜索空间的计算复杂性。
In this paper, we propose a score-based normalizing flow method called DAG-NF to learn dependencies of input observation data. Inspired by Grad-CAM in computer vision, we use jacobian matrix of output on input as causal relationships and this method can be generalized to any neural networks especially for flow-based generative neural networks such as Masked Autoregressive Flow(MAF) and Continuous Normalizing Flow(CNF) which compute the log likelihood loss and divergence of distribution of input data and target distribution. This method extends NOTEARS which enforces a important acylicity constraint on continuous adjacency matrix of graph nodes and significantly reduce the computational complexity of search space of graph.
