Wavelet-based non-parametric HMM's: theory and applications

  title={Wavelet-based non-parametric HMM's: theory and applications},
  author={Laurent Couvreur and Christophe Couvreur},
In this paper, we propose a new algorithm for non-parametric estimation of hidden Markov models (HMM's). The algorithm is based on a \wavelet-shrinkage" density estimator for the state-conditional probability density functions of the HMM's. It operates in an iterative fashion, similar to the EM re-estimation formulae used for maximum likelihood estimation of parametric HMM's. We apply the resulting algorithm to simple examples and show its convergence. The performance of the proposed method is… CONTINUE READING


Publications referenced by this paper.
Showing 1-10 of 13 references

Formulae for mean integrated squared error of nonlinear wavelet-based density estimators

  • P. Hall, P. Patil
  • Ann. Stat., vol. 23,
  • 1995
Highly Influential
4 Excerpts

Density Estimation for Statistics and Data

  • B. W. Silverman
  • 1986
Highly Influential
5 Excerpts

Statistical Modeling by Wavelets

  • B. Vidakovic
  • 1999
1 Excerpt

Xei, \A comparative study of speech detection methods

  • F. S. Van Gerven
  • Proc. Eurospeech,
  • 1997
1 Excerpt

Unsupervised and non-parametric Bayesian classi er for HOS speaker independent HM- M based isolated word speech recognition systems

  • M. Zribi
  • Proc. 8th IEEE Workshop on SSAP,
  • 1996
1 Excerpt

Multiresolution approximations and wavelets orthonormal bases of L2(R)

  • S. G. Mallat
  • Trans. American Math. Soc.,
  • 1989
1 Excerpt

Multiresolution approximations and wavelets orthonormal bases of ¤ Gp C  E ”

  • B. Vidakovic
  • Trans . American Math . Soc .
  • 1989

Similar Papers

Loading similar papers…