×

Gram-Schmidt-Fisher scoring algorithm for parameter orthogonalization in MLE. (English) Zbl 1426.62084

Summary: The estimation of parameters is a key component in statistical modelling and inference. However, parametrization of certain likelihood functions could lead to highly correlated estimates, causing numerical problems, mathematical complexities and difficulty in estimation or erroneous interpretation and subsequently inference. In statistical estimation, the concept of orthogonalization is familiar as a simplifying technique that allows parameters to be estimated independently and thus separates information from each other. We introduce a Fisher scoring iterative process that incorporates the Gram-Schmidt orthogonalization technique for maximum likelihood estimation. A finite mixture model for correlated binary data is used to illustrate the implementation of the method with discussion of application to oesophageal cancer data.

MSC:

62F10 Point estimation
62P10 Applications of statistics to biology and medical sciences; meta analysis

Software:

CORRDAT; MULTIMAX
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] Amari, S. I., Differential geometry of curved exponential families- curvatures and information loss, Annals of Statistics, 10, 357-85 (1982) · Zbl 0507.62026
[2] Amari, S. I., Differential geometry in statistics (1985), New York, NY: Springer Verlag, New York, NY
[3] Bonney, G. E.; Kissling, G. E., Gram-Schmidt orthogonalization of multinormal variates: Applications in genetics, Biometrical Journal, 28, 417-425 (1986) · Zbl 0605.62064
[4] Bonney, G. E., Some new results on regressive models in family studies, Proceedings of the Biometrics Section, American Statistical Association, 177-182 (1995)
[5] Bonney, G. E., Regression analysis of disposition to correlated binary outcomes (1998), Philadelphia, PA: Fox Chase Cancer Center, Philadelphia, PA
[6] Bonney, G. E., Regression analysis of disposition to correlated binary outcomes (1998)
[7] Bonney, G. E.; Kwagyan, J.; Apprey, V., MULTIMAX-A computer package for MULTI-objective MAXimization with applications in genetics and epidemiology, The American Journal of Human Genetics, 61, 447 (1997)
[8] Bonney, G. E.; Kwagyan, J.; Slater, E.; Slifker, M., CORRDAT-A computer package for CORRelated DATa, The American Journal of Human Genetics, 61, A194 (1997)
[9] Cox, D. R.; Reid, N., Parameter orthogonality and approximate conditional inference, Journal of the Royal Statistical Society: Series B, 49, 1-39 (1987) · Zbl 0616.62006
[10] Cox, D. R.; Reid, N., On the stability of maximum likelihood estimators of orthogonal parameters, Canada Journal of Statistics, 17, 229-233 (1989) · Zbl 0679.62017
[11] Clayton, D. G., Gram-Schmidt orthogonalization, Journal of the Royal Statistical Society: Series C (Applied Statistics), 20, 335-338 (1971)
[12] Godambe, V. P., Orthogonality of estimating functions and nuisance parameters, Biometrika, 78, 143-151 (1991)
[13] Godambe, V. P.; Thompson, M. E., An extension of quasi-likelihood estimation (with discussion), Journal of Statistical Planning and Inference, 22, 137-72 (1989) · Zbl 0681.62036
[14] Hills, S. E., Parameter orthogonality and approximate conditional inference [Discussion], Journal of the Royal Statistical Society: Series B, 49, 1-39 (1987)
[15] Hurlimann, W., On parameter orthogonality of the mean, Statistical Papers, 33, 69-74 (1992) · Zbl 0741.62026
[16] Kalbliesch, J. D.; Sprott, D. A., Journal of the Royal Statistical Society. Series B (Methodological), 32, Application of likelihood methods to models involving large number of parameters, 175-208 (1970) · Zbl 0205.45903
[17] Kwagyan, J., Further Investigation of the disposition Model for correlated binary outcomes (2001), Philadelphia, PA: Department of Statistics, Temple University, Philadelphia, PA
[18] Kwagyan, J.; Apprey, V.; Bonney, G. E., Parameter orthogonalization in genetic analysis, Genetic Epidemiology, 21, 163 (2001)
[19] Liang, K. Y.; Zeger, S. L., Longitudinal data analysis using generalized linear models, Biometrika, 73, 13-22 (1986) · Zbl 0595.62110
[20] Prentice, R. L., Correlated binary regression with covariates specific to each binary observation, Biometrics, 44, 1088-1048 (1988) · Zbl 0715.62145
[21] Ross, G. J. S., The efficient use of function minimization in non-linear maximum-likelihood estimation, Applied Statistics, 19, 205-221 (1970)
[22] Willmot, G. E., Parameter orthogonality for a family of discrete distributions, The Journal of the Acoustical Society of America, 83, 517-521 (1988)
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.