What is granular computing? Surprisingly, this book does not provide a precise mathematical definition, but instead gives us 52 papers that aim to illustrate the concepts of granular computing from many different angles. Several intuitive definitions, however, are attempted throughout the book. For example, in his foreword, Zadeh describes granular computing as a mode of computation where the objects of computation are not values of variables, but information about values of variables. According to Kreinovich, granular computing deals with probabilistic, interval, or fuzzy uncertainty, and attempts a combination of precise tools that are able to handle all types of information granules or abstractions. As such, it closely matches human abstraction and problem solving.

The book is divided into three parts. The first part, about fundamentals and methodology, contains 21 papers that cover the background of the main contributing technologies encapsulated by granular computing, namely: interval analysis, fuzzy sets, and rough sets. It also contains several highly representative algorithms of these technologies. The second part, “Hybrid Methods and Models of Granular Computing,” has 13 papers and considers a variety of symbiotic developments of information granules, such as interval-valued fuzzy sets, type-2 fuzzy sets, and shadowed sets. The last part of the book contains 18 papers and considers several applications and case studies.

The book is mainly targeted at established communities in computational intelligence, pattern recognition, machine learning, fuzzy sets, neural networks, system modeling, and operations research. At the same time, it is intended to serve as a reference book for graduate students and senior undergraduate students. This ambitious goal leads to the main problem I had with this book: it tries to be too many things, for too many people, at the same time. As a result, the book loses focus and, at worst, appears to be a somewhat disconnected collection of papers on granular computing and its tools. This book is useful for readers who are already familiar with the area and are looking for a comprehensive collection of work that has been done in the field. Other readers, I am afraid, may be daunted by the book’s sheer size and ambition. For these readers, the book could have benefited from a more rigorous, structured, less fuzzy approach, focusing on fewer applications. I am not sure that in its current format, the book succeeds in bringing granular computing as a unifying theory into the mainstream of all the application areas it mentions.