The aim of this book is to present a survey of the many ways in which the statistical package GLIM may be used to model and analyze stochastic processes. Its emphasis is on using GLIM interactively to apply statistical techniques, and examples are drawn from a wide range of applications including medicine, biology, and the social sciences. It is based on the author's many years of teaching courses along these lines to both undergraduate and graduate students. The author assumes that readers have a reasonably strong background in statistics such as might be gained from undergraduate courses and that they are also familiar with the basic workings of GLIM. Topics covered include: the analysis of survival data, regression and fitting distributions, time series analysis (including both the time and frequency domains), repeated measurements, and generalized linear models.

Ten years ago Bill Gale of AT&T Bell Laboratories was primary organizer of the first Workshop on Artificial Intelligence and Statistics. In the early days of the Workshop series it seemed clear that researchers in AI and statistics had common interests, though with different emphases, goals, and vocabularies. In learning and model selection, for example, a historical goal of AI to build autonomous agents probably contributed to a focus on parameter-free learning systems, which relied little on an external analyst's assumptions about the data. This seemed at odds with statistical strategy, which stemmed from a view that model selection methods were tools to augment, not replace, the abilities of a human analyst. Thus, statisticians have traditionally spent considerably more time exploiting prior information of the environment to model data and exploratory data analysis methods tailored to their assumptions. In statistics, special emphasis is placed on model checking, making extensive use of residual analysis, because all models are 'wrong', but some are better than others. It is increasingly recognized that AI researchers and/or AI programs can exploit the same kind of statistical strategies to good effect. Often AI researchers and statisticians emphasized different aspects of what in retrospect we might now regard as the same overriding tasks.

Tessellations are subdivisions of d-dimensional space into non-overlapping "cells". Voronoi tessellations are produced by first considering a set of points (known as nuclei) in d-space, and then defining cells as the set of points which are closest to each nuclei. A random Voronoi tessellation is produced by supposing that the location of each nuclei is determined by some random process. They provide models for many natural phenomena as diverse as the growth of crystals, the territories of animals, the development of regional market areas, and in subjects such as computational geometry and astrophysics. This volume provides an introduction to random Voronoi tessellations by presenting a survey of the main known results and the directions in which research is proceeding. Throughout the volume, mathematical and rigorous proofs are given making this essentially a self-contained account in which no background knowledge of the subject is assumed.

This book sumarizes recent theoretical and practical developments. The generation and the assessment of pseudo- and quasi-random point sets is one of the basic tasks of applied mathematics and statistics, with implications for Monte Carlo methods, stochastic simulation, and applied statistics. They are also of strong theoretical interest, with applications to algebraic geometry, metric number theory, probability theory, and cryptology.

Most texts on the design of experiments focus on the analysis of experimental data, not on the creation of the design. Graphical Methods for Experimental Design presents a strategic view of the planning of experiments, and provides a number of graphical tools that are useful for justifying the effort required for experimentation, identifying variables and candidate statistical models, selecting the set of run conditions and for assessing the quality of the design. In addition, the graphical framework for creating fractional factorial designs is used to present experimental results in a way that is easier to understand than a set of model coefficients. The text merely assumes a basic knowledge of statistics and matrices, while many of the graphical techniques are accessible without any knowledge of statistical models, requiring only some familiarity with the plotting of functions and with the concept of projection from elementary mechanical drawing.

Robust statistics and the design of experiments are two of the fastest growing fields in contemporary statistics. Up to now, there has been very little overlap between these fields. This is the first book to link these two areas by studying the influence of the design on the efficiency and robustness of robust estimators and tests. The classical approaches of experimental design and robust statistics are introduced before the areas are linked, and the author shows that robust statistical procedures profit by an appropriate choice of the design and that efficient designs for a robust statistical analysis are more applicable.

The need to understand and predict the processes that influence the Earth's atmosphere is one of the grand scientific challenges for the next century. This volume is a series of case studies and review chapters that cover many of the recent developments in statistical methodology that are useful for interpreting atmospheric data. L. Mark Berliner is Professor of Statistics at Ohio State University.

This volume presents the published proceedings of the lOth International Workshop on Statistical Modelling, to be held in Innsbruck, Austria from 10 to 14 July, 1995. This workshop marks an important anniversary. The inaugural workshop in this series also took place in Innsbruck in 1986, and brought together a small but enthusiastic group of thirty European statisticians interested in statistical modelling. The workshop arose out of two G LIM conferences in the U. K. in London (1982) and Lancaster (1985), and from a num ber of short courses organised by Murray Aitkin and held at Lancaster in the early 1980s, which attracted many European statisticians interested in Generalised Linear Modelling. The inaugural workshop in Innsbruck con centrated on GLMs and was characterised by a number of features - a friendly and supportive academic atmosphere, tutorial sessions and invited speakers presenting new developments in statistical modelling, and a very well organised social programme. The academic programme allowed plenty of time for presentation and for discussion, and made available copies of all papers beforehand. Over the intervening years, the workshop has grown substantially, and now regularly attracts over 150 participants. The scope of the workshop is now much broader, reflecting the growth in the subject of statistical modelling over ten years. The elements ofthe first workshop, however, are still present, and participants always find the meetings relevant and stimulating.

Monte Carlo methods are numerical methods based on random sampling and quasi-Monte Carlo methods are their deterministic versions. This volume contains the refereed proceedings of the Second International Conference on Monte Carlo and Quasi-Monte Carlo Methods in Scientific Computing which was held at the University of Salzburg (Austria) from July 9--12, 1996. The conference was a forum for recent progress in the theory and the applications of these methods. The topics covered in this volume range from theoretical issues in Monte Carlo and simulation methods, low-discrepancy point sets and sequences, lattice rules, and pseudorandom number generation to applications such as numerical integration, numerical linear algebra, integral equations, binary search, global optimization, computational physics, mathematical finance, and computer graphics. These proceedings will be of interest to graduate students and researchers in Monte Carlo and quasi-Monte Carlo methods, to numerical analysts, and to practitioners of simulation methods.