Combining neural networks and fuzzy systems, this presents neural networks as trainable dynamical systems and develops mechanisms and principles of adaption, self-organization, covergence and global stability. It also includes the new geometric theory of fuzzy sets, systems and associative memories. Kosko shows how to apply fuzzy theory to adaptive control and how to generate structured fuzzy systems with unsupervised neural techniques. The book is designed to be appropriate for courses in engineering, computer science, mathematics, psychology and biology. Neural, fuzzy and neural-fuzzy IBM PC software is included.