Highly Structured Stochastic Systems edited by Peter Green, Nils Hjort and Sylvia Richardson to be published by Oxford University Press (2002?) Book format There will be 15 chapters, each comprising the main speaker's contribution of around 18 to 20 pages and two discussion contributions of 4 or 5 pages each. In addition, the Editors will contribute an introduction of around 10 pages, outlining what HSSS represents and how the various chapters relate to other and to the field in general. There will also be an index. The book will therefore run to between 425 and 475 pages. Provisional contents (order of chapters is not necessarily final) Introduction, Peter Green, Nils Hjort and Sylvia Richardson Part A: Graphical models and causality 1. Some modern applications of graphical models, Steffen Lauritzen 2. Causal influence using influence diagrams: the problem of partial compliance, Philip Dawid 3. Ancestral graph Markov models, Thomas Richardson and Peter Spirtes 4. Causality and graphical models in time series, Rainer Dahlhaus and Michael Eichler Part B: Spatial statistics 5. Spatial models in epidemiological applications, Sylvia Richardson 6. Spatial hierarchical Bayesian models in ecological applications, Antti Penttinen, Fabio Divino and Anne Riiali 7. Image analysis, Merrilee Hurn, Oddvar Husby and Havard Rue Part C: MCMC 8. Linking theory and practice of MCMC, Gareth Roberts 9. Trans-dimensional MCMC, Peter Green 10. Particle filtering methods for dynamic and static Bayesian problems, Carlo Berzuini and Wally Gilks Part D: Biological applications 11. Preventing epidemics in heterogeneous environments, Niels Becker and Sergey Utev 12. Genetic linkage analysis using MCMC techniques, Simon Heath 13. The genealogy of a neutral mutation, Robert Griffiths and Simon Tavare Part E: Beyond parametric assumptions 14. Topics in nonparametric Bayesian statistics, Nils Hjort 15. HSSS model criticism, Tony O'Hagan Index Background to the HSSS programme Highly Structured Stochastic Systems (HSSS) is the name of an initiative that has been running since 1993 with funding from the European Science Foundation. HSSS refers to a field of Statistics that is characterised by building complex stochastic models from simple localised components. Examples are in image analysis (simple localised descriptions of individual pixels as dependent on neighbouring pixels, that allow complex global image analysis), expert systems (simple conditional relationships between parents and children in graphical models, that build into highly complex knowledge representations) and hierarchical models (where each layer of the hierarchy depends on the next through a simple structure). The HSSS initiative was stimulated by the recognition that these three areas, that had previously been studied independently, had many common features. Since 1993, HSSS has brought together researchers in these and other areas, to explore models and computational tools of common interest. Numerous workshops have been held on particular topics, as well as two international conferences. The end of the HSSS initiative will then be marked by this book, which presents the current state of the field in a comprehensive but accessible form.