A Parallel Framework for Fuzzy Membrane Clustering Based on P Systems and Improved PSO

  • Chengfang Zhang
  • Zhen Yue
  • Jie Jin
  • Dan YanEmail author
Conference paper
Part of the Advances in Intelligent Systems and Computing book series (AISC, volume 856)


The traditional Fuzzy c-means (FCM) clustering algorithm is sensitive to initial seeds and noise. A novel fuzzy clustering algorithm is proposed by Peng Hong, which is named as Fuzzy-MC. However, due to the limitation of the serial architecture of current computer, the parallel and distributed computing characteristics of P system was not able to be exhibited. Therefore, Fuzzy-MC algorithm increases its computing time. To reduce the computing time, this paper proposes a GPU-based parallel Fuzzy-MC algorithm. In the parallel algorithm, block layer in CUDA programming model is used to represent the cells, while threads are used to realize the evolution and communication of the objects. Two artificial data sets and four real-life data sets from the UCI data sets are chosen to compare parallel version and the corresponding serial version. The computing time and clustering performance are used to demonstrate the effectiveness of the proposed algorithm. Compared with the Fuzzy-MC algorithm, parallel Fuzzy-MC algorithm not only ensures the clustering performance but also reduce its computing time.


Fuzzy-MC Improved PSO GPU P systems CUDA 



The authors first sincerely thank the editors and anonymous reviewers for their constructive comments and suggestions. The authors would also like to thank Prof. H Peng from XiHua University (China). This work is supported by the Key research base of philosophy and Social Sciences in Sichuan province and key research base of Humanities and Social Sciences in Sichuan colleges and Universities (No. SHZLQN1701).


  1. 1.
    Sun, J.-G.: Clustering algorithms research. J. Softw. 19(1), 48–61 (2008)CrossRefGoogle Scholar
  2. 2.
    Li, Y.-C., Yin, P.-P.: A FCM-based thread partitioning algorithm for speculative multithreading. Chin. J. Comput. 37(3), 580–592 (2014)MathSciNetGoogle Scholar
  3. 3.
    Guo, H., Liu, L., Zhao, Y., Xu, F.: Sonar image segmentation based on MAR and FCM clustering. Chin. J. Sci. Instrum. 10(34), 2322–2327 (2013)Google Scholar
  4. 4.
    Bezdek, J.C.: Detection and characterization of cluster substructure 1. Linear structure fuzzy c-lines. SIAM J. Appl. Math. 40(2), 339–357 (1981)MathSciNetCrossRefGoogle Scholar
  5. 5.
    Bhuyan, J.N.: Genetic algorithm for clustering with an ordered representation. In: ICGA, pp. 408–415 (1991)Google Scholar
  6. 6.
    Niknam, T., Olamaie, J., Amiri, B.: A hybrid evolutionary algorithm based on ACO and SA for cluster analysis. J. Appl. Sci. 8(15), 1695–2702 (2008)CrossRefGoogle Scholar
  7. 7.
    Omran, M.G., Engelbrecht, A.P.: Self-adaptive differential evolution methods for unsupervised image classification. In: Cybernetics and Intelligent Systems, pp. 1–6 (2006)Google Scholar
  8. 8.
    Hebboul, A., Hachouf, F., Boulemnadjel, A.: A new incremental neural network for simultaneous clustering and classification. Neurocomputing 169(2), 89–99 (2015)CrossRefGoogle Scholar
  9. 9.
    Peng, H., Zhang, J., Jiang, Y.: DE-MC: a membrane clustering algorithm based on differential evolution mechanism. Rom. J. Inf. Sci. Technol. 17(1), 76–88 (2014)Google Scholar
  10. 10.
    Peng, H., Wang, J., Pérez-Jiménez, M.J., Riscos-Núñez, A.: An unsupervised learning algorithm for membrane computing. Inf. Sci. 304(C), 80–91 (2015)CrossRefGoogle Scholar
  11. 11.
    Jin, J., Liu, H., Wang, F., Peng, H.: Parallel implementation of P systems for data clustering on GPU. Bio Inspired Comput. Theor. Appl. 562, 200–211 (2015)CrossRefGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.Sichuan Police CollegeLuzhouChina
  2. 2.Information Security CenterBeijing University of Posts and TelecommunicationsBeijingChina
  3. 3.Chengdu Company, Space Star Technology Co., LTD.ChengduChina

Personalised recommendations