Article
Full entry |
PDF
(0.4 MB)
Feedback
PDF
(0.4 MB)
Feedback
Keywords:
exponential families; information divergence; Fisher information; global information; Akaike criterion; model selection
exponential families; information divergence; Fisher information; global information; Akaike criterion; model selection
Summary:
The concept of global statistical information in the classical statistical experiment with independent exponentially distributed samples is investigated. Explicit formulas are evaluated for common exponential families. It is shown that the generalized likelihood ratio test procedure of model selection can be replaced by a generalized information procedure. Simulations in a classical regression model are used to compare this procedure with that based on the Akaike criterion.
References:
[1] Akaike, H.: Information theory and an extension of the maximum likelihood principle. Proceedings of the Second International Symposium on Information Theory, B. N. Petrov et al. (eds.), Akademiai Kiado, Budapest, 1973, pp. 267–281. MR 0483125 | Zbl 0283.62006
[2] Bauer, P., Pötscher B. M. and Hackl P.: Model selection by multiple test procedures. Statistics 19 (1988), 39–44. DOI 10.1080/02331888808802068 | MR 0921623
[3] Berlinet, A. and Francq, Ch.: Identification of a univariate ARMA model. Comp. Statist. 9 (1994), 117–133. MR 1280755
[4] Brown, L. D.: Fundamentals of Statistical Exponential Families. Inst. of Mathem. Statist, Hayword, California, 1986. MR 0882001 | Zbl 0685.62002
[5] Chernoff, H.: On the distribution of likelihood ratio. Ann. Math. Statist. 25 (1954), 573–578. DOI 10.1214/aoms/1177728725 | MR 0065087
[6] Kullback, S.: Information Theory and Statistics. Wiley, New York, 1959. MR 0103557 | Zbl 0088.10406
[7] Lehmann, E. L.: Testing Statistical Hypotheses. 2nd edition, Wiley, New York, 1986. MR 0852406 | Zbl 0608.62020
[8] Liese, F. and Vajda, J.: Convex Statistical Distances. Teubner, Leipzing, 1987. MR 0926905
[9] Nikolov, V.: Regression and Autoregression Models of Signals and their Recognition by Neural Nets. Diploma Theses. Faculty of Physical and Nuclear Engineering, Czech Tech. University, Prague, 1996. (Czech)
[10] Nishii, R.: Maximum likelihood principle and model selection when the true model is unspecified. Journal of Multivariate Analysis 27 (1988), 392–403. DOI 10.1016/0047-259X(88)90137-6 | MR 0970962 | Zbl 0684.62026
[11] Pötscher, B. M.: Order estimation in Arma–models by Lagrangian multiplier test. Ann. Statist. 11 (1983), 872–885. DOI 10.1214/aos/1176346253 | MR 0707937
[12] Rissanen, J.: Modelling by shortest data description. Automatica 14 (1978), 465–471. DOI 10.1016/0005-1098(78)90005-5
[13] Rockefellar, R. T.: Convex Analysis. Princeton University Press, Princeton, 1970.
[14] Ronchetti, E.: Robust model selection. In Transactions of the Twelfth Prague Conference on Information Theory, ..., J. Á. Víšek, and P. Lachout (eds.), Academy of Sciences of the Czech Republic, Prague, 1994, pp. 200–202.
[15] Rydén, T.: Estimating the order of hidden Markov models. Statistics 26 (1995), 345–354. DOI 10.1080/02331889508802501 | MR 1365683
[16] Sahamoto, Y., Ishiguro, M. and Kitagawa, G.: Akaike Information Criterion Statistics. Reidel, Dordrecht, 1986.
[17] Schwartz, G.: Estimating the dimension of a model. Annals of Statistics 6 (1978), 461–464. DOI 10.1214/aos/1176344136 | MR 0468014
[18] Spanier, J. and Oldham, K. B.: An Atlas of Functions. Springer, Berlin, 1987.
[19] Speed, T. P. and Yu, B.: Model collection and prediction: Normal regression. Ann. Inst. Statist. Math. 45 (1993), 35–54. DOI 10.1007/BF00773667 | MR 1220289
[20] Vajda, I.: Global statistical information, likelihood ratio tests and maximum likelihood estimators. Kybernetika (submitted) (1997).
[21] Vieu, P.: Order choice of nonlinear autoregressive models. Statistics 26 (1995), 307–328. DOI 10.1080/02331889508802499 | MR 1365681
Search
Partner of
