A Decentralized Technique for Robust Probabilistic Mixture Modelling of a Distributed Data Set

  • Ali El Attar
  • Antoine Pigeau
  • Marc Gelgon
Conference paper
Part of the Studies in Computational Intelligence book series (SCI, volume 382)

Abstract

This paper deals with a machine learning task, namely probability density estimation, in the case data is composed of subsets hosted on nodes of a distributed system. Focusing on mixture models and assuming a set of local probability distribution estimates, we demonstrate how it is possible to combining local estimates in a dynamic, robust and decentralized fashion, through gossiping a global probabilistic model over the data set. Experiments are reported to illustrate the proposal.

Keywords

Mixture Model Gaussian Mixture Model Outlier Model Probability Density Estimation Gossip Protocol 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Bishop, C.M.: Pattern Recognition and Machine Learning. Springer, New York (2006)MATHGoogle Scholar
  2. 2.
    Kermarrec, A.-M., van Steen, M.: Gossiping in distributed systems. Operating Systems Review 41(5), 2–7 (2007)CrossRefGoogle Scholar
  3. 3.
    Gu, D.: Distributed EM algorithm for Gaussian Mixtures in Sensor Networks. IEEE Transactions on Neural Networks 19(7), 1154–1166 (2008)CrossRefGoogle Scholar
  4. 4.
    Kowalczyk, W., Vlassis, N.A.: Newscast EM. In: NIPS, pp. 713–720. MIT Press, Cambridge (2004)Google Scholar
  5. 5.
    Chandola, V., Banerjee, A., Kumar, V.: Anomaly detection: A survey. ACM Comput. Surv. 41(3), 1–58 (2009)CrossRefGoogle Scholar
  6. 6.
    Rousseeuw, P.J., Driessen, K.V.: A fast algorithm for the minimum covariance determinant estimator. Technometrics 41(3), 212–223 (1999)CrossRefGoogle Scholar
  7. 7.
    Jelasity, M., Voulgaris, S., Guerraoui, R., Kermarrec, A.-M., van Steen, M.: Gossip-based peer sampling. ACM Trans. Comput. Syst. 25(3), 8 (2007)CrossRefGoogle Scholar
  8. 8.
    Goldberger, J., Roweis, S.T.: Hierarchical clustering of a mixture model. In: NIPS, pp. 505–512. MIT Press, Cambridge (2004)Google Scholar
  9. 9.
    Hershey, J.R., Olsen, P.A.: Approximating the Kullback Leibler divergence between Gaussian mixture models. In: IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2007, vol. 4, pp. IV-317–IV-320 (2007)Google Scholar
  10. 10.
    Arthur, D., Vassilvitskii, S.: k-means++: the advantages of careful seeding. In: SODA 2007: Proceedings of the Eighteenth Annual ACM-SIAM Symposium on Discrete Algorithms, pp. 1027–1035. Society for Industrial and Applied Mathematics, PA (2007)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2011

Authors and Affiliations

  • Ali El Attar
    • 1
  • Antoine Pigeau
    • 1
  • Marc Gelgon
    • 1
  1. 1.LINA (UMR CNRS 6241)Université de NantesNantesFrance

Personalised recommendations