Cross-Entropy Directed Embedding of Network Data
Cross-Entropy Directed Embedding of Network Data
复制标题
网络数据的交叉熵定向嵌入
DOI：
--
复制
发表时间：
2003
期刊：
影响因子：
--
通讯作者：
N. Ueda
中科院分区：
文献类型：
--
作者：
Takeshi Yamada;Kazumi Saito;N. Ueda
我们提出了一种新的方法来嵌入由网络表示的数据到低维欧氏空间。与现有的方法不同，所提出的方法试图最小化的能量函数的基础上的交叉熵之间的理想和嵌入的节点配置，而不直接利用节点之间的成对距离。我们还提出了一个自然的标准来有效地评估嵌入式网络布局的节点连接性如何保持。实验结果表明，该方法提供了更好的布局比一些著名的嵌入方法所产生的建议的标准。我们相信，我们的方法产生了一个大规模网络的自然嵌入，适合通过在二维或三维欧氏空间中手动浏览进行分析。
We present a novel approach to embedding data represented by a network into a low-dimensional Euclidean space. Unlike existing methods, the proposed method attempts to minimize an energy function based on the cross-entropy between desirable and embedded node configurations without directly utilizing pairwise distances between nodes. We also propose a natural criterion to effectively evaluate an embedded network layout in terms of how well node connectivities are preserved. Experimental results show that the proposed method provides better layouts than those produced by some of the well-known embedding methods in terms of the proposed criterion. We believe that our method produces a natural embedding of a large-scale network suitable for analyzing by manual browsing in a two- or three-dimensional Euclidean space.
影响因子：
30.8
作者：
Shen-Orr, SS;Milo, R;Alon, U
通讯作者：
Alon, U