The Prague Symposia on Asymptotic Statistics represent a twenty years' tradi tion of contacts between Czech mathematical statisticians and the conference partic ipants. Both, as the organizers hope, return from the Symposia to their work with fresh ideas and new information. The Fifth Prague Symposium was held from September 4 to September 9, 1993 at the Faculty of Mathematics and Physics, Charles University. It was sponsored by the Bernoulli Society for Mathematical Statistics and Probability, the Czech Statistical the Czech Society of Actuaries, Ceska Pojistovna-Insurance and Reinsur Society, ance Corporation, and the IFIP WG 7. 7. Asymptotic Statistics, a prolific source of methodological concepts, dominated the program of the Symposium. Special sessions were devoted to Mathematics of Insurance and Finance and to Stochastic Programming. The papers presented at the Symposium are published in two parts. Part 1 is . . Part 2 is Number 3, Volume 30 (1994) of the journal Kybernetika, this volume comprising the papers of the authors who were not able to meet the early editorial deadline. The editors of the Proceedings would like to express their sincere thanks to the authors for valuable contributions, to the reviewers for prompt and careful reading the papers, to J. Antoch for his advice with technical part of the Proceedings. Finally they also express their appreciation of the kind cooperation with the Publishing House Physica-Verlag and the journal Kybernetika in bringing out the volumes. Part of the Proceedings was typeset by AN(S-TEX, the macrosystem of the Ame rican Mathematical Society.
Inhaltsverzeichnis
Invited papers. - Procedures for the detection of multiple changes in series of independent observations. - Probing for information in two-stage stochastic programming and the associated consistency. - Outliers and switches in time series. - Fréchet differentiability and robust estimation. - Stein predictors and prediction regions. - Perpetuities and random equations. - On recent developments in the theory of set-indexed processes. - Regression rank scores scale statistics and studentization in linear models. - On an argmax-distribution connected to the Poisson process. - Asymptotic normality of regression M-estimators: Hadamard differentiability approaches. - Contributed papers. - Asymptotic theory for regression quantile estimators in the heteroscedastic regression model. - Nonnegative moving-average models. - Biased samples from a search and the asymptotic behaviour of Bayesian models. - A modification of least squares with high efficiency and high breakdown point in linear regression. - Significance of differences of estimates. - Adaptiveness in time series models. - Kernel estimators of integrated squared density derivatives in non-smooth cases. - Curve selection: A nonparametric approach. - Complete convergence. - Tests for heteroscedasticity based on regression quan-tiles and regression rank scores. - Parameter estimation by projecting on structural statistical models. - Some remarks on the joint estimation of the index and the scale parameter for stable processes. - Asymptotic behaviour of the error probabilities in the pseudo-likelihood ratio test for Gibbs-Markov distributions. - Detection of change in variance. - Shrinkage of maximum likelihood estimator of multivariate location. - Almost sure invariance principles for V- and U-statistics based on weakly dependent random variables. -On stability in two-stage stochastic nonlinear programming. - Spread inequality and efficiency of first and second order. - Confidence intervals for regression quantiles. - Spatial quantiles and their Bahadur-Kiefer representations. - Asymptotic expansions of error probabilities for tests. - An application of complex commutation functions. - One-sided deviations of a random walk without moment assumptions. - Asymptotic behaviour of one-step M-estimators in contaminated non-linear models. - Conditional rank tests for the two-sample problem with partially observed data. - Finiteness and continuity of differential entropy. - Characterizations of discrete probability distributions by the existence of regular conditional distributions respectively continutity from below of inner probability measures. - Diagnostics of regression model: test of goodness of fit. - On a multistage stochastic linear program. - Change point and jump estimates in an AMOC renewal model. - Conditions for consistency of MLE s. - Improving maximum quasi-likelihood estimators.
