• Mathematics
  • Published 1996

Local models and Gaussian mixture models for statistical data processing

@inproceedings{Kambhatla1996LocalMA,
  title={Local models and Gaussian mixture models for statistical data processing},
  author={Nandakishore Kambhatla},
  year={1996}
}
In this dissertation, we present local linear models for dimension reduction and Gaussian mixture models for classification and regression. When the data has different structure in different parts of the input space, fitting once global model can be slow and inaccurate. Simple learning models can quickly learn the structure of the data in small (local) regions. Thus, local learning techniques can offer us faster and more accurate model fitting. Gaussian mixture models form a soft local model of… CONTINUE READING

Citations

Publications citing this paper.
SHOWING 1-10 OF 29 CITATIONS

Dimension Reduction by Local Principal Component Analysis

VIEW 5 EXCERPTS
CITES BACKGROUND & RESULTS
HIGHLY INFLUENCED

Distortion function and clustering for local linear models

VIEW 7 EXCERPTS
CITES BACKGROUND
HIGHLY INFLUENCED

Gaussian Mixture Model Based Coding of Speech and Audio

VIEW 6 EXCERPTS
CITES BACKGROUND & METHODS
HIGHLY INFLUENCED

A candidate solutions generator based on mixed strategy for non-rigid object extraction

Automatic Transition Detection of Segmented Motion Clips Using PCA-based GMM Method

VIEW 1 EXCERPT
CITES BACKGROUND

References

Publications referenced by this paper.
SHOWING 1-10 OF 104 REFERENCES

Principal components analysis of images via back propagation

  • G. W. Cottrell
  • 1988
VIEW 4 EXCERPTS
HIGHLY INFLUENTIAL

Asymptotically optimal block quantization

  • Allen Gersho
  • Mathematics, Computer Science
  • IEEE Trans. Information Theory
  • 1979
VIEW 4 EXCERPTS
HIGHLY INFLUENTIAL

Asymptotic Theory for Principal Component Analysis

VIEW 6 EXCERPTS
HIGHLY INFLUENTIAL

A tutorial on hidden Markov models and selected applications

  • R R.abiner.L.
  • 1989
VIEW 2 EXCERPTS
HIGHLY INFLUENTIAL

Subspace methods of pattern recognition

VIEW 3 EXCERPTS
HIGHLY INFLUENTIAL