Fast Incremental von Neumann Graph Entropy Computation: Theory, Algorithm, and Applications

被引:0
作者
Chen, Pin-Yu [1 ]
Wu, Lingfei [1 ]
Liu, Sijia [1 ]
Rajapakse, Indika [2 ]
机构
[1] IBM Res, Yorktown Hts, NY 10598 USA
[2] Univ Michigan, Ann Arbor, MI 48109 USA
来源
INTERNATIONAL CONFERENCE ON MACHINE LEARNING, VOL 97 | 2019年 / 97卷
关键词
NETWORKS; LAPLACIAN; SPECTRA;
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
The von Neumann graph entropy (VNGE) facilitates measurement of information divergence and distance between graphs in a graph sequence. It has been successfully applied to various learning tasks driven by network-based data. While effective, VNGE is computationally demanding as it requires the full eigenspectrum of the graph Laplacian matrix. In this paper, we propose a new computational framework, Fast Incremental von Neumann Graph EntRopy (FINGER), which approaches VNGE with a performance guarantee. FINGER reduces the cubic complexity of VNGE to linear complexity in the number of nodes and edges, and thus enables online computation based on incremental graph changes. We also show asymptotic equivalence of FINGER to the exact VNGE, and derive its approximation error bounds. Based on FINGER, we propose efficient algorithms for computing Jensen-Shannon distance between graphs. Our experimental results on different random graph models demonstrate the computational efficiency and the asymptotic equivalence of FINGER. In addition, we apply FINGER to two real-world applications and one synthesized anomaly detection dataset, and corroborate its superior performance over seven baseline graph similarity methods.
引用
收藏
页数:11
相关论文
共 63 条
[1]   Graph based anomaly detection and description: a survey [J].
Akoglu, Leman ;
Tong, Hanghang ;
Koutra, Danai .
DATA MINING AND KNOWLEDGE DISCOVERY, 2015, 29 (03) :626-688
[2]   Shannon and von Neumann entropy of random networks with heterogeneous expected degree [J].
Anand, Kartik ;
Bianconi, Ginestra ;
Severini, Simone .
PHYSICAL REVIEW E, 2011, 83 (03)
[3]   Entropy measures for networks: Toward an information theory of complex topologies [J].
Anand, Kartik ;
Bianconi, Ginestra .
PHYSICAL REVIEW E, 2009, 80 (04)
[4]  
Anderson J., 1985, Linear Multilinear Algebra, V18, P141, DOI DOI 10.1080/03081088508817681
[5]  
[Anonymous], 1990, MATRIX ANAL
[6]  
[Anonymous], 1995, Combinatorial Optimization
[7]  
[Anonymous], 2000, Templates for the Solution of Algebraic Eigenvalue Problems: A Practical Guide. Ed. by, DOI DOI 10.1137/1.9780898719581
[8]  
[Anonymous], 2015, P 28 INT C NEURAL IN
[9]  
[Anonymous], 2014, ADV NEURAL INFORM PR
[10]   Depth-based complexity traces of graphs [J].
Bai, Lu ;
Hancock, Edwin R. .
PATTERN RECOGNITION, 2014, 47 (03) :1172-1186