5,457 Hits in 4.7 sec

Subgraph Networks with Application to Structural Feature Space Expansion [article]

Qi Xuan, Jinhuan Wang, Minghao Zhao, Junkun Yuan, Chenbo Fu, Zhongyuan Ruan, Guanrong Chen
2019 arXiv   pre-print
Furthermore, these SGNs are used to expand the structural feature space of the underlying network, beneficial for network classification.  ...  In this paper, the concept of subgraph network (SGN) is introduced and then applied to network models, with algorithms designed for constructing the 1st-order and 2nd-order SGNs, which can be easily extended  ...  These networks together may provide more comprehensive structural information for subsequent applications. In this paper, the focus is on its application to network classification.  ... 
arXiv:1903.09022v2 fatcat:ko6wdvy325dwjk4xkck4vct6vy


Lin Li, Kefeng Fan, Zhiyong Zhang, Zhengmin Xia
2016 Neural Network World  
Community structure implies some features in various real-world networks, and these features can help us to analysis structural and functional properties in the complex system.  ...  To solve this problem, we improve the classic community detection algorithm with Principal Component Analysis(PCA) mapping and local expansion k-means.  ...  We map network nodes into low-dimension space with PCA, and then based on the topology feature of the community structure, the local expansion strategy is proposed to choose reasonable initial seeds.  ... 
doi:10.14311/nnw.2016.26.034 fatcat:vosaqc4acjfqdfhjeb3hx4iep4

Ripple Walk Training: A Subgraph-based training framework for Large and Deep Graph Neural Network [article]

Jiyang Bai, Yuxiang Ren, Jiawei Zhang
2021 arXiv   pre-print
To deal with these problems, in this paper, we propose a general subgraph-based training framework, namely Ripple Walk Training (RWT), for deep and large graph neural networks.  ...  Such problems attribute to the data structures of the graph itself or the designing of the multi-layers GNNs framework, and can lead to low training efficiency and high space complexity.  ...  INTRODUCTION Graph neural networks (GNNs) have achieved outstanding performance in graph-structured data based applications, such as knowledge graphs [28] , social medias [22] , and protein interface  ... 
arXiv:2002.07206v3 fatcat:f4bn5e5wfzgnti53zepxs54nlm

Overlapping decomposition for causal graphical modeling

Lei Han, Guojie Song, Gao Cong, Kunqing Xie
2012 Proceedings of the 18th ACM SIGKDD international conference on Knowledge discovery and data mining - KDD '12  
Based on this theory, a greedy expansion algorithm is proposed to generate the overlapping subgraphs.  ...  Causal graphical models are developed to detect the dependence relationships between random variables and provide intuitive explanations for the relationships in complex systems.  ...  to constrain the structure of the subgraphs.  ... 
doi:10.1145/2339530.2339551 dblp:conf/kdd/HanSCX12 fatcat:tu2zvgssefdqdnf2apfhjc62i4

SURREAL: Subgraph Robust Representation Learning

Saba A. Al-Sayouri, Danai Koutra, Evangelos E. Papalexakis, Sarah S. Lam
2019 Applied Network Science  
Representation learning algorithms aim to preserve local and global network structure by identifying node neighborhoods.  ...  In this paper, we propose SURREAL, a novel, stable graph embedding algorithmic framework that leverages "spatio-electric" (SE) subgraphs: it learns graph representations using the analogy of graphs with  ...  Acknowledgements Not applicable.  ... 
doi:10.1007/s41109-019-0160-1 fatcat:jhiluwpebrfnlk22kyqhxjk4wy

Brain network analysis

Xiangnan Kong, Philip S. Yu
2014 SIGKDD Explorations  
The network structure can be very noisy and uncertain. Therefore, innovative methods are required for brain network analysis. Many research e↵orts have been devoted to this area.  ...  They have achieved great success in various applications, such as brain network extraction, graph mining , neuroimaging data analysis.  ...  Dual Tensor Kernel (DuSK): In order to preserve the tensor structure in both original space and feature space, we can use CP factorizations to define the tensor kernels.  ... 
doi:10.1145/2641190.2641196 fatcat:l7gnbu3zufbvjmw67ptxycyoyu

MODA: An efficient algorithm for network motif discovery in biological networks

Saeed Omidi, Falk Schreiber, Ali Masoudi-Nejad
2009 Genes & Genetic Systems  
Many indicators have been proposed to assess the global features of networks.  ...  While most of the algorithms rely on induced subgraphs as motifs of the networks, MODA is able to extract both induced and non-induced subgraphs simultaneously.  ...  As a result, for an empirical study of theories about large scale features and local structures such as network motifs, it is highly desirable to have efficient tools.  ... 
doi:10.1266/ggs.84.385 pmid:20154426 fatcat:a4hay75difhhvdioxclbc7ixt4

RECS: Robust Graph Embedding Using Connection Subgraphs [article]

Saba A. Al-Sayouri, Danai Koutra, Evangelos E. Papalexakis, Sarah S. Lam
2018 arXiv   pre-print
Representation learning algorithms aim to preserve local and global network structure by identifying node neighborhood notions.  ...  RECS learns graph representations using connection subgraphs by employing the analogy of graphs with electrical circuits.  ...  Assumption 2 -Symmetry in feature space. The source node u and any node w in its refined neighborhood N R (u), have a symmetrical impact on each other in the continuous feature space.  ... 
arXiv:1805.01509v3 fatcat:qc2de3khaff2dfekyxhb7znry4

A survey of frequent subgraph mining algorithms

Chuntao Jiang, Frans Coenen, Michele Zito
2012 Knowledge engineering review (Print)  
The research goals are directed at: (i) effective mechanisms for generating candidate subgraphs (without generating duplicates) and (ii) how best to process the generated candidate subgraphs so as to identify  ...  This paper presents a survey of current research in the field of frequent subgraph mining, and proposed solutions to address the main research issues.  ...  Some unordered tree FTM algorithms are directed at specific applications and can use features of these applications to enhance the efficiency of the operation of the algorithm.  ... 
doi:10.1017/s0269888912000331 fatcat:pxye65ayvzgevplfpkjhwissn4

Subgraphs of Interest Social Networks for Diffusion Dynamics Prediction

Valentina Y. Guleva, Polina O. Andreeva, Danila Vaganov
2021 Entropy  
We explore different types of social networks, demonstrating high structural variability, and aim to extract and see their minimal building blocks, which are able to reproduce supergraph structural and  ...  dynamical properties, so as to be appropriate for diffusion prediction for the whole graph on the base of its small subgraph.  ...  The most similar subgraphs to the initial networks were obtained (sample size was >20) by Community structure expansion, random walk, non-back tracking random walk samplers, and Metropolis-Hasting.  ... 
doi:10.3390/e23040492 pmid:33924216 pmcid:PMC8074582 fatcat:jbegiouohfd2jmtrf5vw2ef2gm

Graph Complexity from the Jensen-Shannon Divergence [chapter]

Lu Bai, Edwin R. Hancock
2012 Lecture Notes in Computer Science  
From the centroid vertex, a family of centroid expansion subgraphs of the graph with increasing layers are constructed.  ...  In this paper we aim to characterize graphs in terms of structural complexities.  ...  Since the L layer centroid expansion subgraph possesses the full structures of any K layer centroid expansion subgraphs, the union graph is the L layer centroid expansion subgraph.  ... 
doi:10.1007/978-3-642-34166-3_9 fatcat:s3qzqwvrcjhczbolnmmcyim7xu

Application of Spectral Clustering Methods in Pipeline Systems Graph Models

Gul'naz I. Galimova, Il'nur D. Galimyanov, Dinar T. Yakupov, Vladimir V. Mokshin
2019 Helix  
Decomposition of any graph as a structure with inherent topology meets the criteria for optimality in connectedness and balanced subgraphs with a small number of clusters.  ...  With an increase in the number of sub-areas above a certain value, the probability of appearance of the disconnected subgraphs in the decomposition structure increases.  ...  Acknowledgements The work is performed according to the Russian Government Program of Competitive Growth of Kazan Federal University.  ... 
doi:10.29042/2019-5607-5614 fatcat:3h5d4a3vn5duvowvsqaifsi3zq

Sampling Subgraph Network with Application to Graph Classification [article]

Jinhuan Wang and Pengtao Chen and Bin Ma and Jiajun Zhou and Zhongyuan Ruan and Guanrong Chen and Qi Xuan
2021 arXiv   pre-print
We also present a hierarchical feature fusion framework to integrate the structural features of diverse sampling SGNs, so as to improve the performance of graph classification.  ...  Recently, a subgraph network (SGN) model is proposed to study the potential structure among motifs, and it was found that the integration of SGN can enhance a series of graph classification methods.  ...  The overall framework of S 2 GN construction for structural feature space expansion is shown in Fig. 4 .  ... 
arXiv:2102.05272v1 fatcat:i7rp24qx4vgwvdxxfqynbhqzo4

Pseudo Cold Start Link Prediction with Multiple Sources in Social Networks [chapter]

Liang Ge, Aidong Zhang
2012 Proceedings of the 2012 SIAM International Conference on Data Mining  
In this work, we introduce the pseudo cold start link prediction with multiple sources as the problem of predicting the structure of a social network when only a small subgraph of the social network is  ...  We propose a two-phase supervised method: the first phase generates an efficient feature selection scheme to find the best feature from multiple sources that is used for predicting the structure in the  ...  The regularization first mapped the points in the social network to a Hilbert space and mandated the mapping to respect the known subgraph and a regularizer.  ... 
doi:10.1137/1.9781611972825.66 dblp:conf/sdm/GeZ12 fatcat:vxv4jbm2bzdbniqpdjv34uw2zi

Discovery of Large Disjoint Motif in Biological Network using Dynamic Expansion Tree [article]

Sabyasachi Patra, Anjali Mohapatra
2018 bioRxiv   pre-print
Identification of such network motifs leads to many important applications, such as: understanding the modularity and the large-scale structure of biological networks, classification of networks into super-families  ...  In this algorithm embeddings corresponding to child node of expansion tree are obtained from the embeddings of parent node, either by adding a vertex with time complexity O(n) or by adding an edge with  ...  Embeddings of the subgraph in the target network are computed along with the growth of expansion tree.  ... 
doi:10.1101/308254 fatcat:5ihdh3esh5hmpflec5dcpot3sm
« Previous Showing results 1 — 15 out of 5,457 results