Push Ranking Learning Algorithm on graphs

摘要：

Data representations with graphs are increasingly used in ontology similarity computation, biomedicine, web design and other fields. The goal of graph learning algorithm is not only for classification or regression of vertices, but also for the ranking of vertices. These kinds of algorithms lay stress on the ranking accuracy of the top ranked vertices, and this is the criterion used when judging whether a ranking function is good or bad. This paper introduces the model for push ranking algorithm on graphs, and choosing a q order hinge loss function to propose a new algorithm. Experimental results show the effectiveness of the proposed new algorithms.

关键词： graph firefighter problem ranking push ranking hinge loss function

作者: Yaya Wang Wei Gao Yungang Zhang Yun Gao

作者单位: Department of Information Engineering, Binzhou Polytechnic, Binzhou, Shandong, China Department of Information, Yunnan Normal University, Kunming, Yunnan, China Department of Mathematic Department of Information, Yunnan Normal University, Kunming, Yunnan, China Department of Computer S Department of Editorial, Yunnan Normal University, Kunming, Yunnan, China

会议类型: 国际会议

会议名称: 2010 International Conference on Circuit and Signal Processing(2010年电路与信号处理国际会议 ICCSP 2010)

会议地点: 上海

会议语种:英文

页码: 368-371

在线出版日期: 2010-12-25（万方平台首次上网日期，不代表论文的发表时间）