NMF and PCA as Applied to Gearbox Fault Data

Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 9375)

Abstract

Both Non-negative matrix factorization (NMF) and Principal component analysis (PCA) are data reduction methods. Both of them act as approximation methods permitting to represent data by lower rank matrices. The two methods differ by their criteria how to obtain the approximation. We show that the main assumption of PCA demanding orthogonal principal components leads to a higher rank approximation as that established by NMF working without that assumption. This can be seen when analyzing a data matrix obtained from vibration signals emitted by a healthy and a faulty gearbox. To our knowledge this fact has not been clearly stated so far and no real example supporting our observation has been shown explicitly.

Keywords

Non-negative matrix factorization Principal component analysis Matrix approximation Optimization with constraints Gearbox fault detection 

References

  1. 1.
    Bartkowiak, A.M., Zimroz, R.: Probabilistic principal components, how this works. In: Saeed, K., Homenda, E. (eds.) CISIM 2015, IFIP LNCS 9339 (in print) (2015)Google Scholar
  2. 2.
    Berry, M.W., et al.: Algorithms and applications for approximate nonnegative matrix factorization. Comput. Stat. Data Anal. 52, 155–173 (2007)MathSciNetCrossRefMATHGoogle Scholar
  3. 3.
    Cichocki, A., Zdunek, R., Phan, A.H., Amari, Sh: Nonnegative Matrix and Tensor Factorizations. Applications to Exploratory Multi-way Data Analysis and Blind Source Separation. Wiley, Chichester (2009)Google Scholar
  4. 4.
    Fevotte, C., Bertin, N., Durrieu, J.-L.: Nonnegative matrix factorization with the Itakura-Saito divergence. With application to music analysis. Neural Comput. 21(3), 793–830 (2009)CrossRefMATHGoogle Scholar
  5. 5.
    Gillis, N.: The why and how of nonnegative matrix factorization. In: Suykens, J.A.K., et al. (eds.) Regularization, Optimization, Kernels and Support Machines, pp. 3–39. Chapman & Hall/CRC, London (2014)Google Scholar
  6. 6.
    Guillamet, D., Vitria, J., Schiele, B.: Introducing a weighted non-negative matrix factorization for image classification. Pattern Recogn. Lett. 24, 2447–2454 (2003)CrossRefMATHGoogle Scholar
  7. 7.
    Hoyer, P.O.: Non-negative matrix factorization with sparseness constraints. J. Mach. Learn. Res. 5, 1457–1469 (2004)MathSciNetMATHGoogle Scholar
  8. 8.
    Jolliffe, I.T.: Principal Component Analysis, 2nd edn. Springer, New York (2002)MATHGoogle Scholar
  9. 9.
    Lee, D.D., Seung, H.S.: Learning the parts of objects by nonnegative matrix factorization. Nature 401, 788–791 (1999)CrossRefGoogle Scholar
  10. 10.
    Li, B., et al.: A new feature extraction and selection scheme for hybrid fault diagnosis of gearbox. Expert Syst. Appl. 38, 1000–1009 (2011)Google Scholar
  11. 11.
    Li, X., Fukui, K.: Fisher non-negative matrix factorization with pairwise weighting. In: IAPR Conference on Machine Vision Application, MVA 2007, pp. 380–383. Tokyo, Japan, May 16–18 (2007)Google Scholar
  12. 12.
    Li, S.Z., et al.: Learning spatially localized, parts-based representation. In: Proceesings of IEEE International Conference on Computer Vision and Pattern Recognition, vol. 1, pp. 207–2012. Havaii, December 2001Google Scholar
  13. 13.
    Tsagaris, V., Anastassopoulos, V.: Feature extraction of Meris data. In: Proceedings of the 2004 Envisat & ERS Symposium, Salzburg, Austria, 6–10 September 2004 (ESA SP-572, April 2005), pp. 1–5 (2005)Google Scholar
  14. 14.
    Zimroz, R., Bartkowiak, A.: Two simple multivariate procedures for monitoring planetary gearboxes in non-stationary operating conditions. Mech. Syst. Signal Process. 38, 237–247 (2013)CrossRefGoogle Scholar
  15. 15.
    Zimroz, R., Bartkowiak, A.: Investigation on spectral structure of gearbox vibration signals by principal component analysis for condition monitoring purposes. J. Phys. Conf. Ser. 305(1), 012075 (2011)CrossRefGoogle Scholar
  16. 16.
    Zdunek, R.: Trust-region algorithm for nonnegative matrix factorization with alpha- and beta-divergences. In: Pinz, A., Pock, T., Bischof, H., Leberl, F. (eds.) DAGM and OAGM 2012. LNCS, vol. 7476, pp. 226–235. Springer, Heidelberg (2012) CrossRefGoogle Scholar
  17. 17.
    Zurada, J.M., Ensari, T., Asi, E.H., Chorowski, J.: Nonnegative matrix factorization and its application to pattern recognition and text mining. In: Proceedings of the 13th FedCSIS Conference on Computer Science and Information Systems, Cracow, pp. 11–16 (2013)Google Scholar

Copyright information

© Springer International Publishing Switzerland 2015

Authors and Affiliations

  1. 1.Institute of Computer ScienceWroclaw UniversityWroclawPoland
  2. 2.Diagnostics and Vibro-Acoustics Science LaboratoryWroclaw University of TechnologyWroclawPoland

Personalised recommendations