Simultaneous Non-gaussian Data Clustering, Feature Selection and Outliers Rejection

  • Nizar Bouguila
  • Djemel Ziou
  • Sabri Boutemedjet
Part of the Lecture Notes in Computer Science book series (LNCS, volume 6744)


A method for simultaneous non-Gaussian data clustering, feature selection and outliers rejection is proposed in this paper. The proposed approach is based on finite generalized Dirichlet mixture models learned within a framework including expectation-maximization updates for model parameters estimation and minimum message length criterion for model selection. Through a challenging application involving texture images discrimination, it is demonstrated that the developed procedure performs effectively in avoiding outliers and selecting relevant features.


  1. 1.
    Fayyad, U., Piatetsky-Shapiro, G., Smyth, P.: Knowledge Discovery and Data Mining: Towards a Unifying Framework. In: Proc. of the Annual ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD), pp. 82–88 (1996)Google Scholar
  2. 2.
    Law, M.H.C., Figueiredo, M.A.T., Jain, A.K.: Simultaneous Feature Selection and Clustering Using Mixture Models. IEEE Transactions on Pattern Analysis and Machine Intelligence 26(9), 1154–1166 (2004)CrossRefGoogle Scholar
  3. 3.
    Boutemedjet, S., Bouguila, N., Ziou, D.: A Hybrid Feature Extraction Selection Approach for High-Dimensional Non-Gaussian Data Clustering. IEEE Transactions on Pattern Analysis and Machine Intelligence 31(8), 1429–1443 (2009)CrossRefGoogle Scholar
  4. 4.
    Bouguila, N., Ziou, D.: High-Dimensional Unsupervised Selection and Estimation of a Finite Generalized Dirichlet Mixture Model Based on Minimum Message Length. IEEE Transactions on Pattern Analysis and Machine Intelligence 29(10), 1716–1731 (2007)CrossRefGoogle Scholar
  5. 5.
    Bouguila, N., Ziou, D.: A New Approach for High-Dimensional Unsupervised Learning: Applications to Image Restoration. In: Pal, S.K., Bandyopadhyay, S., Biswas, S. (eds.) PReMI 2005. LNCS, vol. 3776, pp. 200–205. Springer, Heidelberg (2005)CrossRefGoogle Scholar
  6. 6.
    Williams, C.K.I., Titsias, M.K.: Greedy Learning of Multiple Objects in Images using Robust Statistics and Factorial Learning. Neural Computation 16(5), 1039–1062 (2003)CrossRefzbMATHGoogle Scholar
  7. 7.
    Sudderth, E.B., Torralba, A., Freeman, W.T., Wilsky, A.S.: Depth from Familiar Objects: A Hierarchical Model for 3D Scenes. In: Proc. of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 2410–2417 (2006)Google Scholar
  8. 8.
    Luettgen, M.R., Willsky, A.S.: Likelihood Calculation for a Class of Multiscale Stochastic Models, with Application to Texture Discrimination. IEEE Transactions on Image Processing 4(2), 194–207 (1995)CrossRefGoogle Scholar
  9. 9.
    Ojala, T., Valkealahti, K., Oja, E., Pietikäinen, M.: Texture Discrimination with Multidimensional Distributions of Signed Gray-Level Differences. Pattern Recognition 34, 727–739 (2001)CrossRefzbMATHGoogle Scholar
  10. 10.
    Ojala, T., Mäenpää, T., Viertola, J., Kyllönen, J., Pietikäinen, M.: Empirical Evaluation of MPEG-7 Texture Descriptors with a Large-Scale Experiment. In: Proc. of the 2nd International Workshop on Texture Analysis and Synthesis, pp. 99–102 (2002)Google Scholar
  11. 11.
    Varma, A., Zisserman, A.: Texture classification: are filter banks necessary? In. In: Proc. of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 691–698 (2003)Google Scholar
  12. 12.
    Hayman, E., Caputo, B., Fritz, M., Eklundh, J.-O.: On the significance of real-world conditions for material classification. In: Pajdla, T., Matas, J(G.) (eds.) ECCV 2004. LNCS, vol. 3024, pp. 253–266. Springer, Heidelberg (2004)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2011

Authors and Affiliations

  • Nizar Bouguila
    • 1
  • Djemel Ziou
    • 2
  • Sabri Boutemedjet
    • 2
  1. 1.Concordia Institute for Information Systems EngineeringConcordia UniversityMontrealCanada
  2. 2.Département d’InformatiqueUniversité de SherbrookeCanada

Personalised recommendations