The EM Algorithm used for Gaussian Mixture Modelling and its Initialization

This is an extended version of the paper on EM initialisation, involving a more indepth description and comparison. This file cannot be viewed with ghostview - please use pageview.

[1]  Michael D. Alder,et al.  Adaptive quadratic neural nets , 1991, [Proceedings] 1991 IEEE International Joint Conference on Neural Networks.

[2]  King-Sun Fu,et al.  Syntactic Pattern Recognition And Applications , 1968 .

[3]  D. Rubin,et al.  Maximum likelihood from incomplete data via the EM - algorithm plus discussions on the paper , 1977 .

[4]  A. Cohen,et al.  Finite Mixture Distributions , 1982 .

[5]  A. Morineau,et al.  Multivariate descriptive statistical analysis , 1984 .

[6]  P. McKenzie,et al.  Selecting the optimal number of components for a Gaussian mixture model , 1994, Proceedings of 1994 IEEE International Symposium on Information Theory.

[7]  A. F. Smith,et al.  Statistical analysis of finite mixture distributions , 1986 .

[8]  Michael G. Thomason,et al.  Syntactic Pattern Recognition, An Introduction , 1978, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[9]  Richard A. Johnson,et al.  Applied Multivariate Statistical Analysis , 1983 .

[10]  Line Eikvil,et al.  Statistical classification using a linear mixture of two multinormal probability densities , 1991, Pattern Recognit. Lett..

[11]  M. Alder,et al.  Syntactic pattern recognition by quadratic neural nets. A case study: rail flaw classification , 1993, Proceedings of 1993 International Conference on Neural Networks (IJCNN-93-Nagoya, Japan).