Pattern Analysis and Applications

, Volume 14, Issue 4, pp 349–367 | Cite as

A structurally motivated framework for discriminant analysis

Theoretical Advances

Abstract

Over the last few years, a lot of algorithms for discriminant analysis (DA) have been developed. Although having different motivations, they all inject structure information in data into their own within- and between-class scatters. However, to our best knowledge, there has not been yet a systematical examination about (1) which structure granularities lurk in data; (2) which structure granularities are utilized in scatters of a DA algorithm; (3) whether new DA algorithms can be developed based on existing structure granularities. In this paper, the established so-called structurally motivated (SM) framework for DA and its unified mathematical formulation of the ratio trace exactly answers them. It categorizes these DA algorithms from the viewpoint of constructing scatters based on different-granularity structures in data, identifies their applicable scenarios for different structure types, and provides insights into developing new DA algorithms. Inspired by the insight, we find that cluster granularity lying in the middle of granularity spectrum in SM framework can still be further utilized and exploited. As a result, the three DA algorithms based on the cluster granularity are derived from the SM framework and from the injection of the cluster structure information into the respective within-class, between-class and joint both scatter matrices for the classical MDA, and these corresponding algorithms are, respectively, called as SWDA, SBDA and SWBDA. The injection of cluster structure information makes the proposed three algorithms able to fit relatively complicated data not only more effectively, but also with the regularization technique obtain more projections than the classical MDA, which is very helpful for more effective DA. Moreover, MDA becomes their special case when the cluster numbers of all classes are set to 1. Our experiments on the benchmarks (face and UCI databases) here show that the proposed algorithms yield encouraging results.

Keywords

Discriminant analysis Structure granularity Framework Scatter matrix Classification 

References

  1. 1.
    Duda RO, Hart PE, Stork DG (2001) Pattern Classification, 2nd edn. Wiley, New YorkMATHGoogle Scholar
  2. 2.
    Yan S, Xu D, Zhang B, Zhang H, Yang Q, Lin S (2006) Graph embedding and extension: a general framework for dimensionality reduction. IEEE Trans Pattern Anal Mach Intell 29(1):40–51CrossRefGoogle Scholar
  3. 3.
    Cai D, He XF, Kun Z, Han JW, Bao HJ (2007) Locality sensitive discriminant analysis. In: Proceedings of the international joint conference on artificial intelligence, Hyderabad, India, pp 141–146Google Scholar
  4. 4.
    Manli Z, Martinez AM (2006) Subclass discriminant analysis. IEEE Trans Pattern Anal Mach Intell 28(8):1274–1286CrossRefGoogle Scholar
  5. 5.
    Fukunaga K (1990) Introduction to statistical pattern recognition, 2nd edn. Academic Press, BostonMATHGoogle Scholar
  6. 6.
    Hastie T, Tibshirani R (1996) Discriminant analysis by gaussian mixture. J Roy Stat Soc B 58(1):155–176MathSciNetMATHGoogle Scholar
  7. 7.
    Ye J, Janardan R, Li Q (2005) Two-dimensional linear discriminant analysis. In: Proceedings of advances in neural information processing systems, vol 17, pp 1569–1576Google Scholar
  8. 8.
    Hand D (1982) Kernel discriminant analysis. Research Studies Press, ChichesterMATHGoogle Scholar
  9. 9.
    Harrison RF, Pasupa K (2010) A simple iterative algorithm for parsimonious binary kernel Fisher discrimination. Pattern Anal Appl Theor Adv 13(1):15–22MathSciNetCrossRefGoogle Scholar
  10. 10.
    Shen C, Li H, Brooks MJ (2008) Supervised dimensionality reduction via sequential semidefinite programming. Pattern Recogn 41(12):3644–3652MATHCrossRefGoogle Scholar
  11. 11.
    Yeung DS, Wang D, Wing WYN, Tsang ECC, Zhao X (2007) Structured large margin machines: sensitive to data distributions. Mach Learn 68(2):171–200CrossRefGoogle Scholar
  12. 12.
    Belkin M, Niyogi P (2002) Using manifold structure for partially labeled classification. In: Proceedings of advances in neural information processing systems. http://www.cse.ohio-state.edu/~mbelkin/papers/UMS_NIPS_02.pdf
  13. 13.
    Vapnik VN (1999) The nature of statistical learning theory. Springer, New YorkGoogle Scholar
  14. 14.
    Burges C (1998) A tutorial on support vector machines for pattern recognition. Data Min Knowl Disc 2(2):121–167CrossRefGoogle Scholar
  15. 15.
    Shivaswamy PK, Jebara T (2007) Ellipsoidal kernel machines. In: Proceedings of the 11th artificial intelligence and statistics. San Juan, Puerto Rico, pp 481–489Google Scholar
  16. 16.
    Huang K, Yang H, King I, Lyu MR (2004) Learning large margin classifiers locally and globally. In: Proceedings of the 21st international conference on machine learning. Banff, Canada, pp 51–59Google Scholar
  17. 17.
    Bishop CM (2006) Pattern recognition and machine learning. Springer, BerlinMATHGoogle Scholar
  18. 18.
    Xue H, Chen SC, Yang Q (2008) Structural support vector machine. In: Proceedings of the 15th international symposium on neural networks, LNCS5263, ISNN (1), pp 501–511Google Scholar
  19. 19.
    Boyd S, Vandenberghe L (2003) Convex optimization. Cambridge University Press, CambridgeGoogle Scholar
  20. 20.
    Olvera-López J, Carrasco-Ochoa J, Martínez-Trinidad J (2010) A new fast prototype selection method based on clustering. Pattern Anal Appl Theor Adv 13(2):131–141CrossRefGoogle Scholar
  21. 21.
    Zhu X (2008) Semi-supervised learning literature survey. Computer sciences technical report 1530, University of Wisconsin-MadisonGoogle Scholar
  22. 22.
    Rao CR (2002) Linear statistical inference and its applications, 2nd edn. Wiley, New YorkGoogle Scholar
  23. 23.
    Brand M (2003) Continuous nonlinear dimensionality reduction by kernel eigenmaps. Technical report 2003–21, Mitsubishi Electric Research LabsGoogle Scholar
  24. 24.
    Song Y, Nie F, Zhang C, Xiang S (2008) A unified framework for semi-supervised dimensionality reduction. Pattern Recogn 41(9):2789–2799MATHCrossRefGoogle Scholar
  25. 25.
    Turk M, Pentland A (1991) Face recognition using eigenfaces. In: Proceedings of the computer vision and pattern recognition, pp 586–591Google Scholar
  26. 26.
    Tenenbaum J, Silva V, Langford J (2000) A global geometric framework for nonlinear dimensionality reduction. Science 290(5500):2319–2323CrossRefGoogle Scholar
  27. 27.
    Roweis S, Saul L (2000) Nonlinear discriminant analysis by locally linear embedding. Science 290(5500):2323–2326CrossRefGoogle Scholar
  28. 28.
    Belkin M, Niyogi P (2003) Laplacian eigenmaps for dimensionality redaction and data representation. Neural Comput 15(6):1373–1396MATHCrossRefGoogle Scholar
  29. 29.
    He XF, Niyogi P (2003) Locality preserving projections. In: Proceedings of advances in neural information processing systems, vol 16. Vancouver, Canada, pp 153–160Google Scholar
  30. 30.
    Li H, Jiang T, Zhang K (2006) Efficient and robust feature extraction by maximum margin criterion. IEEE Trans Neural Netw 17(1):157–165CrossRefGoogle Scholar
  31. 31.
    Li Y, Guan C (2007) Joint feature re-extraction and classification using an iterative semi-supervised SVM algorithm. Mach Learn 71(1):33–53CrossRefGoogle Scholar
  32. 32.
    Bengio Y, Delalleau O, Roux NL (2006) The Curse of highly variable functions for local kernel machines. In: Proceedings of advances in neural information processing systems, vol 18, pp 107–114Google Scholar
  33. 33.
    Blake CL, Merz CJ (1998) UCI repository of machine learning databases. University of California, Irvine, Department of Information and Computer SciencesGoogle Scholar
  34. 34.
    Salvador S, Chan P (2004) Determining the number of clusters/segments in hierarchical clustering/segmentation algorithms. In: Proceedings of the 16th IEEE international conference on tools with AI, pp 576–584Google Scholar
  35. 35.
    Ji S, Ye J (2008) A unified framework for generalized linear discriminant analysis. In: Proceedings of IEEE computer society conference on computer vision and pattern recognition, pp 1–7. doi:10.1109/CVPR.2008.4587377
  36. 36.
    Cai D,He X, Han J (2007) Semi-supervised discriminant analysis. In: Proceedings of IEEE international conference on computer vision, Rio de Janeiro, Brazil. doi:10.1109/ICCV.2007.4408856
  37. 37.
    Cai D, He X, Han J (2007) Spectral regression for efficient regularized subspace learning. In: Proceedings of IEEE international conference on computer vision, Rio de Janeiro, Brazil. doi:10.1109/ICCV.2007.4408855
  38. 38.
    Liu J, Chen SC, Tan XY (2008) Fractional order singular value decomposition representation for face recognition. Pattern Recogn 41(1):378–395MathSciNetMATHCrossRefGoogle Scholar
  39. 39.
    Yang T, Kecman V (2010) Face recognition with adaptive local hyperplane algorithm. Pattern Anal Appl Theor Adv 13(1):79–83MathSciNetCrossRefGoogle Scholar
  40. 40.
    Chougdali K, Jedra M, Zahid N (2010) Kernel relevance weighted discriminant analysis for face recognition. Pattern Anal Appl Theor Adv 13(2):213–221MathSciNetCrossRefGoogle Scholar
  41. 41.
    Martinez AM, Kak AC (2001) PCA versus LDA. IEEE Trans Pattern Anal Mach Intell 23(2):228–233CrossRefGoogle Scholar
  42. 42.
    Beyer K, Goldstein J, Ramakrishnan R, Shaft U (1999) When is nearest neighbors meaningful? In: Processing of international conference on database theory, pp 217–235Google Scholar
  43. 43.
    van der Maaten LJP, Postma EO, van den Herik HJ (2009) Dimensionality reduction: a comparative review. Tilburg University Technical Report, TiCC-TR 2009-005Google Scholar
  44. 44.
    Yang B, Chen S (2010) Sample-dependent graph construction with application to dimensionality reduction. Neurocomputing 74(1–3):301–314CrossRefGoogle Scholar

Copyright information

© Springer-Verlag London Limited 2011

Authors and Affiliations

  1. 1.College of Computer Science and TechnologyNanjing University of Aeronautics and AstronauticsNanjingPeople’s Republic of China
  2. 2.Department of Computer ScienceUniversity of VermontBurlingtonUSA

Personalised recommendations