Parameter estimation procedure for the mixtures of normal distributions using discriminant analysis is introduced. Parameter estimation procedure for the mixtures of normal distributions using EM algorithm and using maximum likelihood is reviewed. A simulation study is conducted in comparing five variant methods involving maximum likelihood, EM algorithm, and discriminant analysis.

discriminant analysis, EM algorithm, grid search, maximum likelihood estimate, Newton-Raphson algorithm, normal probability density function.

Received: November 18, 2020; Accepted: December 7, 2020; Published: February 6, 2021

How to cite this article: Mezbahur Rahman and Galkande Iresha Premarathna, Parameter estimation for the mixtures of normal distributions using discriminant analysis, Far East Journal of Theoretical Statistics 61(1) (2021), 1-20. DOI: 10.17654/TS061010001

This Open Access Article is Licensed under Creative Commons Attribution 4.0 International License

References:

[1] E. Acuña, On the modality of finite mixtures of normal densities, Bulletin of the International Statistical Institute, 52nd Session Proceedings, 1999.
[2] K. E. Basford, G. J. McLachlan and M. G. York, Modeling the distribution of stamp paper thickness via finite normal mixtures: the 1872 Hidalgo stamp issue of Mexico revisited, J. Appl. Stat. 24 (1997), 169-180.
[3] I. Eisenberger, Genesis of bimodal distributions, Technometrics 6 (1964), 357-363.
[4] R. V. Hogg, J. W. McKean and A. T. Craig, Introduction to Mathematical Statistics, Prentice Hall, New Jersey, USA, 2005.
[5] M. E. Johnson, Multivariate Statistical Simulation, John Wiley & Sons, New York, USA, 1987.
[6] K. Lange, Statistics and Computing: Numerical Analysis for Statisticians, Springer, New York, USA, 1999.
[7] J. S. Marron and M. P. Wand, Exact mean integrated squared error, Ann. Statist. 20(2) (1992), 712-736.
[8] R. J. Muirhead, Aspects of Multivariate Statistical Theory (Wiley Series in Probability and Statistics), Wiley-Interscience, New York, USA, 1982.
[9] Mezbahur Rahman and Sabiha Mahzabeen, Parameter estimation for the mixtures of normal distributions, Far East Journal of Theoretical Statistics 54(5) (2018), 453-461.
[10] Mezbahur Rahman and Sabiha Mahzabeen, Maximum likelihood parameter estimation for the mixtures of normal distributions, Advances and Applications in Statistics 53(5) (2018), 501-518.
[11] M. Rahman, R. Rahman and L. M. Pearson, Quantiles for finite mixtures of normal distributions, International Journal of Mathematical Education in Science and Technology 37(3) (2006), 352-358.