Local Search with Dynamic-threshold Configuration Checking and Incremental Neighborhood Updating for Maximum k-plex Problem

Peilin Chen
Peilin Chen
Jia Li
Jia Li
Haicheng Chen
Haicheng Chen

national conference on artificial intelligence, 2020.

Cited by: 0|Bibtex|Views48|Links
Keywords:
configuration checkingMaximum Clique Problemlocal search algorithmDynamic-threshold Configuration CheckingDouble-attributes Incremental Neighborhood UpdatingMore(10+)
Weibo:
We evaluate DCCplex on real-world massive graphs from Network Data Repository online, which have recently been used in testing the performance of local search algorithms

Abstract:

The Maximum k-plex Problem is an important combinatorial optimization problem with increasingly wide applications. In this paper, we propose a novel strategy, named Dynamic-threshold Configuration Checking (DCC), to reduce the cycling problem of local search. Due to the complicated neighborhood relations, all the previous local search alg...More

Code:

Data:

0
Introduction
  • In social network analysis, detecting a large cohesive subgraph is a fundamental and extensively studied topic with various applications.
  • In some real-world applications, the networks of interest may be built based on empirical data with noises and faults.
  • In these cases, large cohesive subgraphs hardly appear as ideal cliques.
  • Large cohesive subgraphs hardly appear as ideal cliques
  • To tackle this problem, many clique relaxation models have been proposed (Luce 1950; Mokken 1979; Pattillo et al 2013).
  • The maximum k-plex problem, that is, to find a k-plex of maximum size on a given graph with a given integer k, has received increasing attention from researchers in the fields of social network analysis and data mining (Kondo and Okubo 2012; Xiao et al 2017; Conte et al 2018; Gao et al 2018)
Highlights
  • In social network analysis, detecting a large cohesive subgraph is a fundamental and extensively studied topic with various applications
  • We evaluate DCCplex and FD-TS on DIMACS and BHOSLIB benchmark with a cutoff time of 1000s
  • We evaluate DCCplex on real-world massive graphs from Network Data Repository online (Rossi and Ahmed 2015), which have recently been used in testing the performance of local search algorithms (Rossi et al 2014; Wang, Cai, and Yin 2016; Cai, Lin, and Luo 2017)
  • We have proposed a novel variant of Configuration Checking named Dynamic-threshold Configuration Checking (DCC) and the Double-attribute Neighborhood Update (DINU) scheme for the maximum k-plex problem
  • Based on the DCC strategy and Double-attributes Incremental Neighborhood Updating (DINU) scheme, we develop a local search algorithm DCCplex
  • The experimental result shows that DCCplex achieve good performance across a broad range of problem instances and update the lower bounds on the size of the maximum k-plexes on many hard instances
Results
  • Evaluation on DIMACS and BHOSLIB

    The authors evaluate DCCplex and FD-TS on DIMACS and BHOSLIB benchmark with a cutoff time of 1000s.
  • Within the cutoff time of 1000s, BnB returns a solution and proves its optimality on 76 out of 108 (= 36 × 3) instances.
  • For these 76 instances, DCCplex can achieve the same quality solution within the cutoff time of 100s.
Conclusion
  • The authors have proposed a novel variant of Configuration Checking named DCC and the Double-attribute Neighborhood Update (DINU) scheme for the maximum k-plex problem.
  • Based on the DCC strategy and DINU scheme, the authors develop a local search algorithm DCCplex.
  • The experimental result shows that DCCplex achieve good performance across a broad range of problem instances and update the lower bounds on the size of the maximum k-plexes on many hard instances.
  • SCCplex DCCplex #Better #Better k=2 1(1) 3(10).
Summary
  • Introduction:

    In social network analysis, detecting a large cohesive subgraph is a fundamental and extensively studied topic with various applications.
  • In some real-world applications, the networks of interest may be built based on empirical data with noises and faults.
  • In these cases, large cohesive subgraphs hardly appear as ideal cliques.
  • Large cohesive subgraphs hardly appear as ideal cliques
  • To tackle this problem, many clique relaxation models have been proposed (Luce 1950; Mokken 1979; Pattillo et al 2013).
  • The maximum k-plex problem, that is, to find a k-plex of maximum size on a given graph with a given integer k, has received increasing attention from researchers in the fields of social network analysis and data mining (Kondo and Okubo 2012; Xiao et al 2017; Conte et al 2018; Gao et al 2018)
  • Results:

    Evaluation on DIMACS and BHOSLIB

    The authors evaluate DCCplex and FD-TS on DIMACS and BHOSLIB benchmark with a cutoff time of 1000s.
  • Within the cutoff time of 1000s, BnB returns a solution and proves its optimality on 76 out of 108 (= 36 × 3) instances.
  • For these 76 instances, DCCplex can achieve the same quality solution within the cutoff time of 100s.
  • Conclusion:

    The authors have proposed a novel variant of Configuration Checking named DCC and the Double-attribute Neighborhood Update (DINU) scheme for the maximum k-plex problem.
  • Based on the DCC strategy and DINU scheme, the authors develop a local search algorithm DCCplex.
  • The experimental result shows that DCCplex achieve good performance across a broad range of problem instances and update the lower bounds on the size of the maximum k-plexes on many hard instances.
  • SCCplex DCCplex #Better #Better k=2 1(1) 3(10).
Tables
  • Table1: Evaluation on DIMACS and BHOSLIB with k = 2, 3, 4
  • Table2: Evaluation on Massive Graphs with k = 2, 3, 4
  • Table3: Comparing SCCplex and DCCplex
Download tables as Excel
Funding
  • This paper was supported by the National Natural Science Foundation of China (No 61573386, 61976232), National Key R&D Program of China (No 2018YFC0830600), Guangdong Province Natural Science Foundation (No 2016A030313292, 2017A070706010 (soft science), 2018A030313086), Guangdong Province Science and Technology Plan projects (No 2016B030305007 and No 2017B010110011), Guangzhou Science and Technology Project (No 201804010435)
  • Shaowei Cai was supported by Youth Innovation Promotion Association, Chinese Academy of Sciences (No.2017150)
Reference
  • Balasundaram, B.; Butenko, S.; and Hicks, I. V. 201Clique relaxations in social network analysis: The maximum k-plex problem. Operations Research 59(1):133–142.
    Google ScholarLocate open access versionFindings
  • Battiti, R., and Protasi, M. 2001. Reactive local search for the maximum clique problem. Algorithmica 29(4):610–637.
    Google ScholarLocate open access versionFindings
  • Boginski, V.; Butenko, S.; and Pardalos, P. M. 2006. Mining market data: a network approach. Computers & Operations Research 33(11):3171–3184.
    Google ScholarLocate open access versionFindings
  • Butenko, S., and Wilhelm, W. E. 2006. Clique-detection models in computational biochemistry and genomics. European Journal of Operational Research 173(1):1–17.
    Google ScholarLocate open access versionFindings
  • Cai, S., and Su, K. 2013. Local search for boolean satisfiability with configuration checking and subscore. Artificial Intelligence 204:75–98.
    Google ScholarLocate open access versionFindings
  • Cai, S.; Lin, J.; and Luo, C. 2017. Finding A small vertex cover in massive sparse graphs: Construct, local search, and preprocess. Journal of Artificial Intelligence Research 59:463–494.
    Google ScholarLocate open access versionFindings
  • Cai, S.; Luo, C.; and Su, K. 2015. Improving walksat by effective tie-breaking and efficient implementation. Computer Journal 58(11):2864–2875.
    Google ScholarLocate open access versionFindings
  • Cai, S.; Su, K.; and Sattar, A. 2011. Local search with edge weighting and configuration checking heuristics for minimum vertex cover. Artificial Intelligence 175(9-10):1672– 1696.
    Google ScholarLocate open access versionFindings
  • Conte, A.; De Matteis, T.; De Sensi, D.; Grossi, R.; Marino, A.; and Versari, L. 2018. D2k: Scalable community detection in massive networks via small-diameter k-plexes. In Proceedings of the 24th ACM SIGKDD, 1272–1281.
    Google ScholarLocate open access versionFindings
  • Gao, J.; Chen, J.; Yin, M.; Chen, R.; and Wang, Y. 2018. An exact algorithm for maximum k-plexes in massive graphs. In Proceedings of the 27th International Joint Conference on Artificial Intelligence, 1449–1455.
    Google ScholarLocate open access versionFindings
  • Glover, F., and Laguna, M. 1998. Tabu search. In Handbook of Combinatorial Optimization. 2093–2229.
    Google ScholarFindings
  • Gujjula, K. R.; Seshadrinathan, K. A.; and Meisami, A. 2014. A hybrid metaheuristic for the maximum k-plex problem. In Examining Robustness and Vulnerability of Networked Systems. 83–92.
    Google ScholarFindings
  • Hoos, H. H., and Stutzle, T. 2004. Stochastic local search: Foundations and applications. Elsevier.
    Google ScholarFindings
  • Kondo, K., and Okubo, T. 2012. Structural estimation and interregional labour migration: Evidence from japan. Keio/Kyoto Joint Global COE Discussion Paper Series 2011-040.
    Google ScholarFindings
  • Lakhlef, H. 20A multi-level clustering scheme based on cliques and clusters for wireless sensor networks. Computers & Electrical Engineering 48:436–450.
    Google ScholarLocate open access versionFindings
  • Lin, J.; Cai, S.; Luo, C.; Lin, Q.; and Zhang, H. 2019. Towards more efficient meta-heuristic algorithms for combinatorial test generation. In Proceedings of Joint Meeting of the European Software Engineering Conference and the ACM SIGSOFT Symposium on the Foundations of Software Engineering 2019, 212–222.
    Google ScholarLocate open access versionFindings
  • Luce, R. D. 1950. Connectivity and generalized cliques in sociometric group structure. Psychometrika 15(2):169–190.
    Google ScholarLocate open access versionFindings
  • Luo, C.; Yu, J.; Yu, D.; and Cheng, X. 2015. Distributed algorithms for maximum clique in wireless networks. In Proceedings of the 11th International Conference on Mobile Ad-hoc and Sensor Networks, 222–226.
    Google ScholarLocate open access versionFindings
  • Luo, C.; Su, K.; and Cai, S. 2012. Improving local search for random 3-sat using quantitative configuration checking. In Proceedings of the 20th European Conference on Artificial Intelligence, 570–575. IOS Press.
    Google ScholarLocate open access versionFindings
  • McClosky, B., and Hicks, I. V. 2012. Combinatorial algorithms for the maximum k-plex problem. Journal of combinatorial optimization 23(1):29–49.
    Google ScholarLocate open access versionFindings
  • Miao, Z., and Balasundaram, B. 2017. Approaches for finding cohesive subgroups in large-scale social networks via maximum k-plex detection. Networks 69(4):388–407.
    Google ScholarLocate open access versionFindings
  • Mokken, R. J. 1979.
    Google ScholarFindings
  • Pattillo, J.; Veremyev, A.; Butenko, S.; and Boginski, V. 2013. On the maximum quasi-clique problem. Discrete Applied Mathematics 161(1-2):244–257.
    Google ScholarLocate open access versionFindings
  • Pullan, W. 2007. Protein structure alignment using maximum cliques and local search. In Proceedings of Advances in Artificial Intelligence 2007, 776–780.
    Google ScholarLocate open access versionFindings
  • Rossi, R. A., and Ahmed, N. K. 2015. The network data repository with interactive graph analytics and visualization. In Proceedings of the 29th AAAI Conference on Artificial Intelligence, 4292–4293.
    Google ScholarLocate open access versionFindings
  • Rossi, R. A.; Gleich, D. F.; Gebremedhin, A. H.; and Patwary, M. M. A. 2014. Fast maximum clique algorithms for large graphs. In Proceedings of the 23rd International World Wide Web Conference, 365–366.
    Google ScholarLocate open access versionFindings
  • Wang, Y.; Cai, S.; Chen, J.; and Yin, M. 2018. A fast local search algorithm for minimum weight dominating set problem on massive graphs. In Proceedings of the 27th International Joint Conference on Artificial Intelligence, 1514– 1522.
    Google ScholarLocate open access versionFindings
  • Wang, Y.; Li, C.; Sun, H.; Chen, J.; and Yin, M. 2019. Mlqcc: an improved local search algorithm for the set kcovering problem. International Transactions in Operational Research 26(3):856–887.
    Google ScholarLocate open access versionFindings
  • Wang, Y.; Cai, S.; and Yin, M. 2016. Two efficient local search algorithms for maximum weight clique problem. In Proceedings of the 30th AAAI Conference on Artificial Intelligence, 805–811.
    Google ScholarLocate open access versionFindings
  • Xiao, M.; Lin, W.; Dai, Y.; and Zeng, Y. 2017. A fast algorithm to compute maximum k-plexes in social network analysis. In Proceedings of the 31st AAAI Conference on Artificial Intelligence, 919–925.
    Google ScholarLocate open access versionFindings
  • Xu, K.; Boussemart, F.; Hemery, F.; and Lecoutre, C. 2005. A simple model to generate hard satisfiable instances. arXiv preprint cs/0509032.
    Google ScholarFindings
  • Zhou, Y., and Hao, J.-K. 2017. Frequency-driven tabu search for the maximum s-plex problem. Computers & Operations Research 86:65–78.
    Google ScholarLocate open access versionFindings
Your rating :
0

 

Tags
Comments