TY - JOUR
T1 - Enhance the Performance of Network Computation by a Tunable Weighting Strategy
AU - Li, Hui Jia
AU - Bu, Zhan
AU - Wang, Zhen
AU - Cao, Jie
AU - Shi, Yong
N1 - Publisher Copyright:
© 2017 IEEE.
PY - 2018/6
Y1 - 2018/6
N2 - Networked systems with high computational efficiency are desired in many applications ranging from sociology to engineering. Generally, the performance of the network computation can be enhanced by two ways: rewiring and weighting. In this paper, we proposed a new two-modes weighting strategy based on the concept of communication neighbor graph, which takes use of both the local and global topological properties, e.g., degree centrality, betweenness centrality, and closeness centrality. The weighting strategy includes two modes: In the original mode, it enhances the network synchronizability by increasing the weights of bridge edges; whereas in the inverse version, it increases the significance of community structure by decreasing the weights of bridge edges. The scheme of weighting is controlled by only one parameter, i.e., \alpha, which can be easily performed. We test the effectiveness of our model on a number of artificial benchmark networks as well as real-world ones. To the best of our knowledge, the proposed weighting strategy can outperform the existing methods in improving the performance of network computation.
AB - Networked systems with high computational efficiency are desired in many applications ranging from sociology to engineering. Generally, the performance of the network computation can be enhanced by two ways: rewiring and weighting. In this paper, we proposed a new two-modes weighting strategy based on the concept of communication neighbor graph, which takes use of both the local and global topological properties, e.g., degree centrality, betweenness centrality, and closeness centrality. The weighting strategy includes two modes: In the original mode, it enhances the network synchronizability by increasing the weights of bridge edges; whereas in the inverse version, it increases the significance of community structure by decreasing the weights of bridge edges. The scheme of weighting is controlled by only one parameter, i.e., \alpha, which can be easily performed. We test the effectiveness of our model on a number of artificial benchmark networks as well as real-world ones. To the best of our knowledge, the proposed weighting strategy can outperform the existing methods in improving the performance of network computation.
KW - communication neighbor graph
KW - community structure
KW - Complex networks
KW - synchronizability
KW - weighting strategy
UR - http://www.scopus.com/inward/record.url?scp=85049444422&partnerID=8YFLogxK
U2 - 10.1109/TETCI.2018.2829906
DO - 10.1109/TETCI.2018.2829906
M3 - 文章
AN - SCOPUS:85049444422
SN - 2471-285X
VL - 2
SP - 214
EP - 223
JO - IEEE Transactions on Emerging Topics in Computational Intelligence
JF - IEEE Transactions on Emerging Topics in Computational Intelligence
IS - 3
ER -