Dissertation > Excellent graduate degree dissertation topics show

Abstract

Widely in nature and human society there are a variety of complex systems, complex systems can be described through a complex network. Complex network of research will greatly contribute to the research and development of complex systems is important to understand the structure and function of complex systems. In recent years, the study of complex network is penetrated into many different disciplines, from physics to biology, and in-depth exploration, scientific understanding of the qualitative features of the complex network of quantitative laws and possible applications, has become a complex system or complex scientific research extremely important in a challenging subject. Link prediction is an emerging research directions in the complex network refers to the network nodes and the network structure using known information to predict network exist but have not yet found an unknown link does not exist, but may be formed link. In recent years, the link has forecast its important theoretical value and the potential applications and wide attention, has become one of the hot research field of the study of complex networks. Link prediction research focuses on an undirected network has no right to or weighted network link prediction problem less. This thesis is not entitled to the network link prediction algorithm based on undirected, were developed to network link prediction algorithm and weighted network link prediction improved algorithm. This thesis is divided into four chapters, the first chapter briefly describes the complex network link prediction and its research significance. The second chapter reviews progress in undirected unweighted network link prediction. In the third chapter, we first of 12 for no expansion to the network link prediction algorithms to the network, and establish a basic framework to link prediction algorithm based on local connection information. Then, based on the statistical analysis of the network motif, we construct a generalized Total neighbors Index, also proposed a two-index predicted binding index. 10 real network link prediction algorithm based on these indices created 16 testing and analysis, certain guiding significance for the practical application of the conclusions. In particular, due to the high prediction accuracy and low computational complexity, generalized neighbor index and the combined index is expected to be applied in the actual link information mining. In the fourth chapter, we propose a suitable weighted network link prediction improved algorithm, which has been tested in several real weighted network analysis of the strong, the weak link prediction accuracy, found the weak link in achieving high precision prediction of the link has a more important role than the strong link. Finally, we summarized the paper on possible future research direction prospect.

Related Dissertations

CLC: > SCIENCE AND > Journal of Systems Science > Systems,modern systems theory > Large scale system theory
© 2012 www.DissertationTopic.Net Mobile

Link Perdiction in Directed and Weighted Networks

Abstract

Related Dissertations