点击次数：

论文类型：期刊论文

发表时间：2019-09-01

发表刊物：JOURNAL OF STATISTICAL MECHANICS-THEORY AND EXPERIMENT

收录刊物：SCIE

ISSN号：1742-5468

关键字：network reconstruction; optimization over networks

摘要：This paper considers the problem of link prediction in scale-free networks under partial observation conditions. We present an optimization approach that involves both node similarity and node degree distribution similarity of the original network and the predicted network. This enables our method to take advantage of local as well as global structural information of the networks under consideration, giving more accurate prediction results compared with existing methods based merely on node similarity indices. Experiments in synthetic scale-free networks and the arXiv citation network confirm the availability and effectiveness of our method.