A Comparative Study of Stellar Spectra Analysis with Neural Networks in Transformed Domains

  • Diego Ordóñez
  • Carlos Dafonte
  • Minia Manteiga
  • Bernardino Arcay
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 5788)


The main purpose of the GAIA mission of the European Space Agency (ESA) is to carry out a stereoscopic census of our galaxy and its environment. This task is based on measurements that will provide with unprecedented exactitude information on the astrometry (distance, movements, and spectral energy distribution) of approximately 1% of the objects in the milky way (109 objects). In the case of the brightest objects, essentially stars, spectra with intermediate resolution in the region of the infrared CaII triplet will be measured with a dedicated spectrograph, RVS (Radial Velocity Spectrometer). Stars can be characterized on the basis of their principal atmospheric parameters: effective temperature, gravity, metal content (general abundance of elements other than H and He), and their abundance of alpha elements (elements with Z>22, α), which provide information on the physical environment in which the star was born. The goal of the present work is to study spectral parameterization by means of ANN: it determines the optimal domain for the ANNs performance, and proposes an adequate noise detection and filtering algorithm by considering simulated data (synthetic spectra) in the spectral region of the RVS.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Rodriguez, A., Arcay, B., Dafonte, C., Manteiga, M., Carricajo, I.: An automated knowledge-based analysis and classification of stellar spectra using fuzzy reasoning. Expert Systems with applications 27(2), 237–244 (2004)CrossRefGoogle Scholar
  2. 2.
    Singh, H.P., Gulati, R.K., Gupta, R.: Stellar spectral classification using principal component analysis and artificial neural networks. MNRAS 295, 312–318 (1998)CrossRefGoogle Scholar
  3. 3.
    Hippel, T.V., Allende, C., Sneden, C.: Automated stellar spectral classification and parameterization for the masses. In: The Garrison Festschrift conference proceedings (June 2002)Google Scholar
  4. 4.
    Fiorentin, P., Bailer-Jones, C., Lee, Y., Beers, T., Sivarani, T., Wilhelm, R., Allende, C., Norris, J.: Estimation of stellar atmospheric parameters from sdss/segue spectra. Astronomy and Astrophysics 467, 1373–1387 (2007)CrossRefGoogle Scholar
  5. 5.
    Recio-Blanco, A., de Laverny, P., Plez, B.: Rvs-arb-001. European Space Agency technique note (2005)Google Scholar
  6. 6.
    Cooley, J.W., Tukey, J.: An algorithm for the machine calculation of complex fourier series. Math. Comput. 19, 297–301 (1965)MathSciNetCrossRefzbMATHGoogle Scholar
  7. 7.
    Mallat, S.: A theory for multiresolution signal decomposition: The wavelet representation. Proc. IEEE Trans. on Pattern. Anal. 11(7), 674–693 (1989)CrossRefzbMATHGoogle Scholar
  8. 8.
    Rumelhart, D.E., Hinton, G.E., Williams, R.J.: Learning representations by back-propagating errors. Nature 323, 533–536 (1986)CrossRefGoogle Scholar
  9. 9.
    Leshno, M., Schocken, S.: Multilayer feedforward networks with non-polynomial activation functions can approximate any function. SSRN eLibrary (1993)Google Scholar
  10. 10.
    Recio-Blanco, A., Bijaoui, A., de Laverny, P.: Automated derivation of stellar atmospheric parameters and chemical abundances: the matisse algorithm. Royal Astronomical Society 370(1), 141–150 (2006)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2009

Authors and Affiliations

  • Diego Ordóñez
    • 1
  • Carlos Dafonte
    • 1
  • Minia Manteiga
    • 2
  • Bernardino Arcay
    • 1
  1. 1.Information and Communications Technologies Department, Faculty of Computer ScienceUniversity of La CoruñaA CoruñaSpain
  2. 2.Department of Navigation and Earth SciencesUniversity of A CoruñaA CoruñaSpain

Personalised recommendations