《计算机应用研究》|Application Research of Computers

基于几何邻居的半监督节点分类

Semi-supervised node classification based geometric neighbor

免费全文下载 (已被下载 次)  
获取PDF全文
作者 成天英,王茜,袁丁
机构 重庆大学 计算机学院,重庆 400044
统计 摘要被查看 次,已被下载
文章编号 1001-3695(2020)09-006-2595-05
DOI 10.19734/j.issn.1001-3695.2019.03.0110
摘要 目前基于网络结构的节点分类方法只注重局部网络连接关系。为了能获取更广泛的网络信息,提出一种基于邻居节点结构信息的半监督节点分类算法CBGN。首先,在网络中加入惩罚因子来改进随机游走策略以获取节点的不定长游走序列,这些节点序列被当做句子输入到word2vec模型中,从而将网络结构的潜在信息转换成向量作为节点的特征表示;其次,改进支持向量机算法,结合梯度下降法和坐标下降法来优化参数空间,以对未标记节点进行更准确的分类;最后,在四个标准数据集上与目前较先进的几种方法进行了对比实验。结果表明,CBGN算法提高了分类精度,相比之前已有的方法具有更好的分类效果。
关键词 特征表示; 节点分类; 半监督学习; 随机游走; 网络分析
基金项目
本文URL http://www.arocmag.com/article/01-2020-09-006.html
英文标题 Semi-supervised node classification based geometric neighbor
作者英文名 Cheng Tianying, Wang Qian, Yuan Ding
机构英文名 College of Computer,Chongqing University,Chongqing 400044,China
英文摘要 The existent node classification methods based on network structure only pay attention to local network connection relationship. For obtaining wider network information, this paper developed a semi-supervised node classification algorithm(CBGN) based on geometric neighbor structure information. This algorithm improved random walk strategy with penalty factor in network to obtain arbitrary length node sequence for each node. It input these node sequences into the word2vec model for transforming the potential information into node vectors. CBGN combined gradient descent method and the coordinate descent method to optimize the SVM classification model. This method compared with current methods on four standard datasets. The results verify that the proposed algorithm improves the classification accuracy and has better classification effect.
英文关键词 feature representation; node classification; semi-supervised learning; random walk; network analysis
参考文献 查看稿件参考文献
 
收稿日期 2019/3/19
修回日期 2019/6/6
页码 2595-2599
中图分类号 TP301.6
文献标志码 A