Unsupervised Classification with Non-Gaussian Mixture Models Using ICA

@inproceedings{Lee1998UnsupervisedCW,
  title={Unsupervised Classification with Non-Gaussian Mixture Models Using ICA},
  author={Te-Won Lee and Michael S. Lewicki and Terrence J. Sejnowski},
  booktitle={NIPS},
  year={1998}
}
We present an unsupervised classification algorithm based on an ICA mixture model. The ICA mixture model assumes that the observed data can be categorized into several mutually exclusive data classes in which the components in each class are generated by a linear mixture of independent sources. The algorithm finds the independent sources, the mixing matrix for each class and also computes the class membership probability for each data point. This approach extends the Gaussian mixture model so… CONTINUE READING

References

Publications referenced by this paper.
Showing 1-10 of 18 references

A unifying framework for independent component analysis

  • Lee, T .-W, M. Girolami, A. J. Bell, T. J. Sejnowski
  • International Journal on Mathematical and…
  • 1999
Highly Influential
5 Excerpts

Independent component analysis using an extended infomax algorithm for mixed sub-gaussian and supergaussian sources

  • Lee, T.-W, M. Girolami, T. J. Sejnowski
  • Neural Computation,
  • 1999
Highly Influential
5 Excerpts

Mixture density networks

  • C. Bishop
  • Technical Report,
  • 1994
Highly Influential
3 Excerpts

Inferring sparse , overcomplete image codesusing an e  cient coding framework

  • M. Lewicki, B. Olshausen
  • Advances in Neural Information Processing Systems
  • 1998
2 Excerpts

Learning nonlinear overcomplete represenations for e  cient coding

  • M. Lewicki, T. J. Sejnowski
  • Advances in Neural Information Processing Systems
  • 1998
2 Excerpts

Similar Papers

Loading similar papers…