Parsimonious formulations for low-diameter clusters
Parsimonious formulations for low-diameter clusters
复制标题
低直径簇的简约公式
DOI:
10.1007/s12532-020-00175-6
复制
发表时间:
2020
影响因子:
6.3
通讯作者:
Buchanan, Austin
中科院分区:
文献类型:
--
作者:
Salemi, Hosseinali;Buchanan, Austin
In the analysis of networks, one often searches for tightly knit clusters. One property of a “good” cluster is a small diameter (say, bounded byk), which leads to the concept of ak-club. In this paper, we propose new path-like and cut-like integer programming formulations for detecting these low-diameter subgraphs. They simplify, generalize, and/or dominate several previously existing formulations. Our best-performing formulation uses only node variables (quite unlike previous formulations) and imposes the diameter-at-most-kconstraints via an exponentially large class of cut-like inequalities. A relatively simple implementation of the cut-like formulation easily outperforms previous approaches, solving dozens of instances of the maximumk-club problem in a second or two that would take hours by other formulations. Moreover, the cut-like formulation is more general in the sense that it applies even when distances are not measured in terms of hops. While we consider only thek-club problem in this paper, the proposed techniques may also be useful in other applications where compact solutions are key (e.g., political districting and wildlife reserve design).
登录
查看更多内容
影响因子:
0.8
作者:
Balabhaskar Balasundaram
通讯作者:
Balabhaskar Balasundaram
DOI:
10.1287/opre.2019.1970
发表时间:
2020-06
期刊:
Oper. Res.
影响因子:
--
作者:
J. Walteros;Austin Buchanan
通讯作者:
J. Walteros;Austin Buchanan
DOI:
10.1109/sffcs.1999.814618
发表时间:
1999
期刊:
40th Annual Symposium on Foundations of Computer Science (Cat. No.99CB37039)
影响因子:
--
作者:
R. Lipton;Anastasios Viglas
通讯作者:
Anastasios Viglas
影响因子:
1.6
作者:
Moradi, Esmaeel;Balasundaram, Balabhaskar
通讯作者:
Balasundaram, Balabhaskar
DOI:
10.1287/ijoc.2019.0914
发表时间:
2020-04
期刊:
INFORMS J. Comput.
影响因子:
--
作者:
Hamidreza Validi;Austin Buchanan
通讯作者:
Hamidreza Validi;Austin Buchanan