Skip to main content

Über dieses Buch

Although no-one is, probably, too enthused about the idea, it is a fact that the development of most empirical sciences to a great extend depends of the development of data analysis methods and techniques, which, due to the necessity of applications of computers for that pur­ pose, actually means that it practically depends on the advancements and orientation of computational statistics. This volume contains complete texts of the lectures held during the Summer School on "Computational Aspects of Model Choice" orga­ nized jointly by Charles University, Prague, and International Associa­ tion for Statistical Computing (IASC) on July 1-14, 1991, in Prague. Main aims of the Summer School were to review and analyse some of the recent developments concerning computational aspects of the model choice as well as their theoretical background. The topics covers the problems of the change point detection, robust estimation and its computational aspects, classification using binary trees, stochastic ap­ proximation and optimization including the discussion about available software, computational aspects of graphical model selection and mul­ tiple hypotheses testing. The bridge between these different approaches is formed by the survey paper about statistical applications of artificial intelligence.



Tomáš Havránek

Tomáš Havránek was born in Prague in the family of well known bohemist academician B. Havránek. His carrier started in 1972 just after finishing Charles University and fulfilling military service. The first job has been that of statistician — consultant in the Institute of Microbiology of the Czechoslovak Academy of Sciences. Here he split interests into the two parts, the routine statistical analysis of biological data and his own scientific problems. And on this place he has found a lot of ideas for books, papers and lectures which followed soon.
Jaromír Antoch

Bernd Streitberg

Bernd Streitberg was born on December 18, 1946 in Bad Kissingen and belonged to an old Franconian family. He saw himself as a ‘genuine’ Franconian, but his favorite city was Berlin, where he had lived most of the time since he was a student. His early academic career was marked by the 1968 students’ movement, which greatly influenced life in Berlin and which he enthusiastically supported. Since he was studying at the Institute of Sociology of the Free University and lived in the students’ dormitory in Schlachtensee, he found himself at the center of these events. But his enthusiasm for different philosophies and brilliant argument of any kind precluded attempt to tie him to special political interests.
Jaromír Antoch

Change Point Problem

The main objective of this paper is to survey procedures that one connected with the change point problem and with testing the constancy of regression relationship over a time. We will focus on the case of independent observations. Detailed algorithms are presented for most of the discussed methods.
Jaromír Antoch, Marie Hušková

Robust Estimation in Linear Model and its Computational Aspects

Heuristics of robust estimation is briefly explained and a survey of the most typical robust methods of the point estimation both of the location and scale parameters and in the linear model is presented. Basic notions of robust statistics as well as of some particular topics of robust diagnostics are discussed, too. Testing of submodels and the possibilities of adaptive approach in the estimation of parameters in the linear model are also commented.
Jaromír Antoch, Jan Ámos Víšek

Constructing Prediction Trees from Data: The RECPAM Approach

Growing trees from the data is presented as a general way of solving the prediction problem for an unknown parameter of a distribution. A tree- structured prediction model is proposed and a strategy for building such a model from the data is presented. The strategy comprises three steps:
RECursive partition;
hence its acronym RECPAM. The construction is based on an information measure, the role of which is highlighted. It is shown that virtually all the available tree-growing approaches are particular cases of the general strategy.
Antonio Ciampi

Stochastic Approximation and Optimization

This is a brief introduction to the topic of stochastic approximation and optimization. An example from the field of water management is given.
Václav Dupač

Comparison of the Stochastic Approximation Software

In this short review the available PC software for stochastic optimization based on the recursive estimation is described. Especially, Stochastic quasigradient software SQG-PC and Stochastic nonlinear problem software SNLP are compared. In the last section problems of possible ‘outliers’ in the optimization problems are considered and their influence to the behavior of recursive process. Some hints are given how to overcome this outlier difficulties together with software recommendation.
Pavel Charamza

Models, Algorithms and Software of Stochastic Optimization

The main objective of this paper is to consider stochastic generalization of linear and nonlinear optimization problems, methods of solution and relevant software.
Victor Loskutov

Some Computational Aspects of Graphical Model Selection

Two alternative approaches to graphical model selection — stepwise edge elimination, and a so-called fast method proposed by Edwards and Havránek (1985, 1987) — are described and compared. Some emphasis is given to specific non — numerical computational aspects: in particular, an efficient algorithm for the dual representation problem is described. The model selection methods are applied to a contingency table concerning risk factors for coronary heart disease.
David Edwards

Multiple Hypotheses Testing

The paper is mainly concerned with multiple testing procedures which control a given multiple level α. General concepts for this purpose are the closure test and a modification which is independent of the special structure of hypotheses and tests. We consider improvements of this modification using information about the logical dependences (redundancies) within the system of hypotheses and present an efficient algorithm. Finally, we discuss some problems which are specific for hierarchical systems of hypotheses, e.g. in model search.
Gerhard Hommel, Gudrun Bernhard

Statistical Applications of Artificial Intelligence

Artificial Intelligence (AI) has now provided some effective techniques for formalization of knowledge about goals and actions. These techniques could open new areas of research to statisticians. Experimental systems designed to assist users of statistics have been constructed in experiment design, data analysis technique application, and technique selection. Knowledge formalization has also been used in experimental programs to assist statisticians in doing data analysis and in building consultation systems.
The best-explored application of AI techniques is building consultation systems. Many small systems have been built, but few systems have been offered for sale, and fewer yet have found their way into regular use. It has become apparent that this is a harder problem than was expected, although the existing successful systems suggest that it is still worth research exploration.
Analogies with successful Artificial Intelligence applications in other fields suggest other statistical applications worth exploring.
Opening new areas to research and providing new tools to users would make considerable changes in the use and production of statistical techniques. However, applying currently available AI techniques will lead to more work for statisticians, not less.
This review is an updated version of (Gale, 1987).
William A. Gale
Weitere Informationen