COMPSTAT 1990 - Proceedings in Computational Statistics: 9th Symposium held at Dubrovnik, Yugoslavia, 1990

Although no-one is, probably, too enthused about the idea, it is a fact that the development of most empirical sciences to a great extent depends on the development of data analysis methods and techniques, which, due to the necessity of application of computers for that purpose, actually means that it practically depends on the advancement and orientation of computer statistics. Every other year the International Association for Statistical Computing sponsors the organizition of meetings of individual s professiona77y involved in computational statistics. Since these meetings attract professionals from allover the world, they are a good sample for the estimation of trends in this area which some believe is a statistics proper while others claim it is computer science. It seems, though, that an increasing number of colleagues treat it as an independent scientific or at least technical discipline. This volume contains six invited papers, 41 contributed papers and, finally, two papers which are, formally, software descriptions, but it was agreed by the Program Committee that they should be included in a separate section entitled "Notes about new developments in statistical software", due to their special significance for current trends in computational statistics.

Classification.- Stochastic Algorithms for Clustering.- A Classification Algorithm for Binary Observations Based on Euclidean Representations of Hypergraphs.- Agglomerative Hierarchical Multicriteria Clustering using Decision Rules.- GLIMTREE:RECPAM Trees with the Generalized Linear Model.- Algorithms and statistical software.- BOJA: A Program for Bootstrap and Jackknife.- Prototyping Dynamic Graphics Functions in S.- Programming Languages for Statistical Computation.- Expert systems in statistics.- Statistical Knowledge-Based Systems ― Critical Remarks and Requirements for Approval.- New Approach To GUHA-Method from the Reliability Viewpoint.- Classifying Documents: A Discriminant Analysis and an Expert System Work Together.- Estimation Procedures for Language Context: Poor Estimates are Worse than None.- Knowledge Modelling for Statistical Consultation Systems; Two Empirical Studies.- An Expert System Strategy for Selecting Interesting Results.- Computer Assisted Interpretation of Conditional Independence Graphs.- WAMASTEX ― Heuristic Guidance for Statistical Analysis.- Multivariate data analysis and model building.- On Model Search Methods.- Principal Components Analysis with Respect to Instrumental Variables via Univariate Splines.- Selecting the Best Subset of Variables in Principal Component Analysis.- Interesting Projections of Multidimensional Data by Means of Generalized Principal Component Analyses.- Maximum Likelihood Estimation of Mixed Linear and Multiplicative Models for Contingency Tables Using Distan.- Alternate Forms of Graphical Modelling ― A Comparison.- Exact Significance Testing by the Method of Control Variates.- Testing Collapsibility of Hierarchical Loglinear Models for Contingency Tables.- The Generalised Biplot: Software Potential.- Exploratory Approach and Maximum Likelihood Estimation of Models for Non Symmetrical Analysis of Two-Way Multiple Contingency Tables.- An Orthogonal Procrustes Rotation Procedure for Multiple Correspondence Analysis.- Optimization techniques and nonlinear models.- Optimization in Statistics ― Recent Trends.- Fitting Non Linear Models with Two Components of Error.- Computing for robust statistics.- Some Proposals for Fast HBD Regression.- Robust Selection of Variables in the Discriminant Analysis Based on MVE and MCD Estimators.- Interactively Computing Robust Covariance Matrices.- Sensitivity Analysis in Factor Analysis: Methods and Software.- Influence Functions of Eigenvalues and Eigenvectors in Multidimensional Data Analysis.- Algorithms for Non-Linear Huber Estimation.- Statistics and database management.- Co-Operative Processing ― A Challenge for Statistics and Database Management.- A Structured Language for Modelling Statistical Data.- Time dependent models.- Spectral Analysis of Non-Stationary Time Series.- ARMA Simulator for Testing Methods of Process Diagnostic.- Statistical Inference in an Extremal Markovian Model.- Interpretation of Spectral Estimation Results for Nonstationary Signals.- An Automated Method for Trend Analysis.- Analysis of spatial data.- A Test of Spatial Isotropy.- Computational inference.- Characteristics of Sequential Sampling Plans for Attributes: Algorithms for Exact Computing.- Exact Experimental Designs via Stochastic Optimization for Nonlinear Regression Models.- A Comparison of Algorithms for Combination of Information in Generally Balanced Designs.- A Comparison of Non-Negative Estimators for Ratios of Variance Components.- Optimal Fit in Non-Parametric Modelling via Computationally Intensive Inference.- Notes about new developments in statistical software.- Statistical Models in S.- GLIM4 ― Developments in Model Fitting.- Address list of authors.