Abstract
In this paper we deal with the problem of classifying a p-dimensional random vector into one of two elliptically contoured populations with unknown and distinct mean vectors and a common, but unknown, scale matrix. The classification procedure is based on two-step monotone training samples, one from each population, with the same monotone pattern. Our aim is to extend the classification procedure, which proposed recently by Chung and Han (Ann Ins Stat Math 52:544–556, 2000). This procedure is a linear combination of two discriminant functions, one based on the complete samples and the other on the incomplete samples. The performance of the proposed classification rule is compared with the plug-in method, this means with the classification rule which arises if the unknown parameters are substituted, into the usual classification rule, by their estimators. In order to apply the plug-in method, the MLE of the location parameters and of the common scale matrix of g ≥ 2 elliptically contoured populations are analytically obtained on the basis of two-step monotone training samples.
Similar content being viewed by others
References
Anderson TW (1957) Maximum likelihood estimates for multivariate normal distribution when some observations are missing. J Am Stat Assoc 52:200–203
Anderson TW (2003) An Introduction to multivariate statistical analysis. Wiley, New York
Anderson TW, Fang KT (1990) Inference in multivariate elliptically contoured distributions based on maximum likelihood in statistical inference. In: Fang KT, Anderson TW (eds) Elliptically contoured and related distributions. Allerton Press, New York, pp 201–216
Batsidis A, Zografos K (2006) Statistical inference for location and scale of elliptically contoured models with monotone missing data. J Stat Plan Inference (in press)
Bello AL (1993) Choosing among imputation techniques for incomplete multivariate data: a simulation study. Commun Stat Theory Methods 22:853–877
Bohannon TR, Smith WB (1975) Classification based on incomplete data records. In: ASA proceedings of social statistics section, pp 214–218
Cacoullos T, Koutras M (1985) Minimum-distance discrimination for spherical distributions. In: Matusita K (eds) Statistical theory and data analysis. Elsevier, New York, pp 91–102
Cacoullos T, Koutras M (1997) On the performance of minimum-distance classification rules for Kotz-type elliptical distributions. In: Johnson NL, Balakrishan N (eds) Advances in the theory and practice of statistics A volume in honor of Samuel Kotz. Wiley, New York, pp 209–224
Chan LS, Dunn OJ (1972) The treatment of missing values in discriminant analysis-1, the sampling experiment. J Am Stat Assoc 67:473–477
Chan LS, Dunn OJ (1974) A note on the asymptotical aspect of the treatment of missing values in discriminant analysis. J Am Stat Assoc 69:672–673
Chan LS, Gilman JA, Dunn OJ (1976) Alternative approaches to missing values in discriminant analysis. J Am Stat Assoc 71:842–844
Chung H, Han C (2000) Discriminant analysis when a block of observations is missing. Ann Ins Stat Math 52:544–556
Fang KT, Zhang YT (1990) Generalized multivariate analysis. Science Press Beijing and Springer, Berlin Heidelberg New York
Fang KT, Kotz S, Ng KW (1990) Symmetric multivariate and related distributions. Chapman and Hall, London, New York
Guerrero-Cusumano JS (1998) Measures of dependence for the multivariate t distribution with applications to the stock market. Commun Stat Theory Methods 27:2985–3006
Gupta AK, Nagar DK (2000) Matrix variate distributions. Chapman and Hall, London
Gupta AK, Varga T (1993) Elliptically contoured models in statistics. Kluwer, Dordrecht
Hao J, Krishnamoorthy K (2001) Inferences on a normal covariance matrix and generalized variance with monotone missing data. J Multivariate Anal 78:62–82
Hufnagel G (1988) On estimating missing values in linear discriminant analysis-part I. Biom J 30:69–75
Joarder AH, Ali MM (1997) Estimation of the scale matrix of a multivariate t-model under entropy loss. Metrika 46:21–32
Joarder AH, Singh S (2001) Estimation of the trace of the scaled covariance matrix of a multivariate t-model using known information. Metrika 54:53–58
Kanda T, Fujikoshi Y (1998) Some basic properties of the MLE’S for a multivariate normal distribution with monotone missing data. Am J Math Manage Sci 18:161–190
Kendall MG (1966) Discrimination and Classification. In: Krishnaiah PR (eds) Multivariate analysis. Academic Press, New York
Koutras M (1987) On the performance of the linear discriminant function for s pherical distributions. J Multivarite Anal 22:1–12
Koutras M (1992). Minimum distance discrimination rules and success rates for elliptical normal mixtures. Stat Probab Lett 13:259–268
Little RJA (1978) Consistent regression methods for discriminant analysis with incomplete data. J Am Stat Assoc 73:319–322
Little RJA, Rubin DB (1987) Statistical analysis with missing data. Wiley, New York
Rubin DB (1976) Inference and missing data. Biometrika 63:581–592
Schafer JL (1997) Analysis of incomplete multivariate data. Chapman and Hall, London
Srivastava JN, Zaatar MK (1972) On the maximum likelihood classification rule for incomplete multivariate samples and its admissibility. J Multivariate Analysis 2:115–126
Sutradhar BC (1990) Discrimination of observations into one of two t populations. Biometrics 46:827–835
Sutradhar BC, Ali MM (1986) Estimation of the parameters of regression model with multivariate-t error variable. Commun Stat Theory Methods 15:429–450
Sutradhar BC, Ali, MM (1989) A generalization of the Wishart distribution for the elliptical model and its moments for the multivariate t model. J Multivariate Anal 29:155–162
Twedt DJ, Gill DS (1992) Comparison of algorithm for replacing missing data in discriminant analysis. Commun Stat Theory Methods 21:1567–1578
Wakaki H (1994) Discriminant analysis under elliptical populations. Hiroshima Math J 24: 257–298
Wakaki H (1997) Asymptotic expansion of the joint distribution of sample mean vector and sample covariance matrix from an elliptical population. Hiroshima Math J 27:295–305
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Batsidis, A., Zografos, K. Discrimination of Observations into One of Two Elliptic Populations based on Monotone Training Samples. Metrika 64, 221–241 (2006). https://doi.org/10.1007/s00184-006-0046-y
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00184-006-0046-y
Keywords
- Monotone missing data
- Elliptically contoured distributions
- Estimation
- Discriminant Analysis
- Error rate
- Multivariate t-distribution