This book presents the concepts needed to deal with self-organizing complex systems from a unifying point of view that uses macroscopic data. The various meanings of the concept "information" are discussed and a general formulation of the maximum information (entropy) principle is used. With the aid of results from synergetics, adequate objective constraints for a large class of self-organizing systems are formulated and examples are given from physics, biology, and computer science (pattern recognition by parallel computers). The extensions contained in the second edition show how, based on possibly scarce and noisy data, unbiased guesses about processes of complex systems can be made and the underlying deterministic and random forces determined. This procedure allows probabilistic predictions of processes, with applications to numerous fields ranging from technology through biology and medicine to economy. The relationship to chaos theory is also addressed.