保留低阶和高阶关系的图表示深度学习集成算法

高质量学习图中节点的低维表示是当前的一个研究热点。现有浅模型的方法无法捕捉图结构的非线性关系，图神经网络技术中的图卷积模型会产生过平滑问题。同时，如何确定不同跳数关系在图表示学习中的作用亦是研究中尚需解决的问题。以解决上述问题为目的，提出一个基于T(T>1)个前馈神经网络的深度学习模型，该框架利用深度模型抽取图结构的非线性关系，T个子模型有效地捕获图的局部和全局(高阶)关系信息，并且它们在最终的向量表示中赋予了不同的作用、从而发挥不同跳数关系的优势。在顶点分类和链接预测任务中的实验结果表明，该框架比现有方法具有竞争力，对比基准算法可以获得20%左右的提升。

Preserving low-order and high-order relationships deep learning ensemble algorithm for graph representation

High-quality learning low-dimensional representation of nodes in the graph is a current research hotspot. The existing shallow model methods cannot capture the nonlinear relationship of the graph structure, and the graph convolution model in the graph neural network technology will cause an over-smoothing problem. At the same time, how to determine the role of different hop number relationships in graph representation learning is also a problem that needs to be solved in the research. To solve the above problems, this paper proposed a deep learning model based on T(T>1) feedforward neural networks. The framework used deep learning models to extract the nonlinear relationship of the graph structure, and T sub-models effectively capture the local and global(higher-order) relationship information of the graph, and they gave different roles in the final vector representation to take advantage of different hop relations. Experimental results on vertex classification and link prediction tasks show that the framework is competitive with existing methods, the benchmark algorithm can be improved by about 20%.