No label defined (Q4650356)

For the first edition (1992) see Zbl 0753.60055. As the author mentions, the main changes concern the chapters on probability metrics and spectral gap. Besides, the references have been updated and numerous corrections and small modifications of presentation undertaken. Some sections are based on new results by the author and his colleagues. The fifteen chapters of this book are split into four parts. Each chapter ends up with (a bit too short) sections with references and historical notes. In the first part the author discusses Markov processes on a general state space. The main tool are the rates (or derivatives) of the transition probabilities called \(q\)-functions or \(Q\)-matrices. Using analytical tools (transition kernels, resolvents, Kolmogorov equations) the author addresses the issues of the uniqueness and existence problem for the processes with given \(q\)-functions. A particular attention is devoted to jump processes. Then the author discusses the issues of recurrence, ergodicity, invariant measures, embedded jump process, and birth-death processes. Finally, this part provides a good exposition of the Kantorovich metric and related coupling issues. It should be noted that the author seldom uses the tools based on the infinitesimal generator and martingale methods. Part II begins with the study of reversible and symmetrizable jump processes and Dirichlet forms. Then the author describes main concepts of the field theory (that describes the communicating states and the potential differences corresponding to the allowed transitions), in particular, electric fields and their application to the study of transience for symmetrizable Markov chains and random walks on fractals. Starting from the first principles, the author elaborates the large deviation techniques for Markov chains. This part ends up with the discussion of spectral theory, Dirichlet forms and isoperimetric (Cheeger's) constants. Part III concerns equilibrium particle systems, in particular Gibbs fields on the grid, related existence and uniqueness issues, and the Peierls method. Further topics include the reversible spin process and exclusion process, Yang-Mills lattice field and the related diffusion process. Part IV deals with non-equilibrium particle systems, construction and existence, their stationary distributions, phase transitions and hydrodynamic limits. The Schlögl model of chemical reaction with diffusion is used to illustrate several results throughout the book, especially those concerning non-equilibrium systems. The index is a bit too short for the book of this length. For instance, the index does not include the \(q\)-process, which is one of the principal objects of this book. It is a pity that a list of notations is not present, since many results are formulated using only mathematical symbols without recalling in words their meaning. However, the book is a welcome addition to the Markov processes literature and will be useful to anyone working in this area, especially for specialists in interacting particle systems, Gibbs processes and probabilistic aspects of statistical physics.