Abstract
A frequently occurring problem is to find a probability vector,p∈D, which minimizes theI-divergence between it and a given probability vector π. This is referred to as theI-projection of π ontoD. Darroch and Ratcliff (1972,Ann. Math. Statist.,43, 1470–1480) gave an algorithm whenD is defined by some linear equalities and in this paper, for simplicity of exposition, we propose an iterative procedure whenD is defined by some linear inequalities. We also discuss the relationship betweenI-projection and the maximum likelihood estimation for multinomial distribution. All of the results can be applied to isotonic cone.
Similar content being viewed by others
References
Agresti, A. (1984).Analysis of Ordinal Categorical, Wiley, New York.
Agresti, A. and Coull, B. (1996). Order-restricted test for stratified comparisons of binomial proportions,Biometrics,52, 1103–1111.
Csiszar, I. (1967). Information-type measure of difference of probability distributions and indirect observations,Studia Sci. Math. Hungar.,2, 299–318.
Csiszar, I. (1975).I-divergence geometry of probability distributions and minimization problems,Ann. Probab.,3, 146–159.
Darroch, J. N. and Ratcliff, D. (1972). Generalized iterative scaling for log-linear models,Ann. Math. Statist.,43, 1470–1480.
Dykstra, R. L. (1985). An iterative procedure for obtainingI-projections onto the intersection of conve sets,Ann. Probab.,13, 975–984.
Good, I. J. (1963). Maximum entropy for hypothesis formulation, especially for multidimensional contingency tables,Ann. Math. Statist.,34, 911–934.
Kemperman, J. H. B. (1969). On the optimum rate of transmitting information,Ann. Math. Statist.,40, 2156–2177.
Kullback, S. (1959).Information Theory and Statistics, Wiley, New York.
Kullback, S. (1967). An extension of information-theoretic derivation of certain limit relations for a Markov chain,SIAM J. Control.,5, 51–53.
Lemke, J. H. and Dykstra, R. L. (1984). An algorithm for multinomial maximum likelihood estimation with multiple cone restrictions. Tech. Report, 84-1, Department of Preventive Medicine, University of Iowa.
McDonald, J. W. and Diamond, I. (1983). Fitting generalized linear models with linear inequality parameter constraints,The GLIM Newsletter,8, 29–36.
McDonald, J. W. and Diamond, I. (1990). On the fitting of generalized linear models with nonnegativity parameter constraints,Biometrics,46, 201–206.
Robertson, T., Wright, F. T. and Dykstra, R. L. (1988).Order Restricted Statistical Inference, Wiley, New York.
Author information
Authors and Affiliations
About this article
Cite this article
Gao, W., Shi, NZ. I-projection onto isotonic cones and its applications to maximum likelihood estimation for log-linear models. Ann Inst Stat Math 55, 251–263 (2003). https://doi.org/10.1007/BF02530498
Received:
Revised:
Issue Date:
DOI: https://doi.org/10.1007/BF02530498