This book is based on the premise that the entropy concept, a fundamental element of probability theory as logic, governs all of thermal physics, both equilibrium and nonequilibrium. The variational algorithm of J. Willard Gibbs, dating from the 19th Century and extended considerably over the following 100 years, is shown to be the governing feature over the entire range of thermal phenomena, such that only the nature of the macroscopic constraints changes. Beginning with a short history of the development of the entropy concept by Rudolph Clausius and his predecessors, along with the formalization of classical thermodynamics by Gibbs, the first part of the book describes the quest to uncover the meaning of thermodynamic entropy, which leads to its relationship with probability and information as first envisioned by Ludwig Boltzmann. Recognition of entropy first of all as a fundamental element of probability theory in mid-twentieth Century led to deep insights into both statistical mechanics and thermodynamics, the details of which are presented here in several chapters. The later chapters extend these ideas to nonequilibrium statistical mechanics in an unambiguous manner, thereby exhibiting the overall unifying role of the entropy.

The book aims to give an overview of the previous Sitges Conferences, which have been held during the last 25 years, with special emphasis on topics related to non-equilibrium phenomena. It includes review articles and articles dealing with new trends in the subject, written by scientists who have played an important role in the development of this area. The book is intended as a commemorative edition of the Sitges Conferences. Graduate students of physics and researchers will find this a stimulating account of the development of non-equilibrium statistical mechanics in the last years, covering a wide scope of topics: kinetic theory, hydrodynamics, fluctuation phenomena and stochastic processes, relaxation phenomena, kinetics of phase transitions, growth kinetics, and so on.

All papers were peer reviewed. Bayesian Inference and Maximum Entropy Methods in Science and Engineering provide a framework for analyzing ill-conditioned data. Maximum Entropy is a theoretical method to draw conclusions when little information is available. Bayesian probability theory provides a formalism for scientific reasoning by analyzing noisy or imcomplete data using prior knowledge.

General physics, atomic physics, molecular physics, and solid state physics.