Visualization and optimization (Q2364984)

When the big mainframe computer was turned into a small desktop appliance, the view of the computer as a tool soon reached its climax and disclosed the view as a medium. Due to great progress in bitmapped graphics, this development is strongly connected with the flux of digital imagery, and the tremendous attention that the visual domain has gained over the last decade. Visualization has become the prevalent catchword. Optimization of complex systems with lots of variables and constraints has always been a particularly promising application of computers. The book by Jones sets out to tell researches and practitioners in the operations research domain about the marvels and promises visualization may offer them. The general assumption is that humans' astonishing visual capabilities could be exploited to the benefit of understanding complex phenomena. If visualized, abstract mental models we always have of phenomena could be communicated and used to improve mutual understanding. The book thus attempts to link the precise, mathematical world of well-defined optimization problems and algorithms with that of ill-defined visualizations. It is divided into three parts. Part I describes some general prerequisites for the use of visualization techniques in optimization. Part II is a collection of specific techniques for the visual presentation of abstract models. Part III (almost half of the book) applies those techniques to optimization problems. The three parts are further subdivided into 6, 4, and 9 chapters respectively. There is an author index, a subject index, and bibliographies are printed at the end of each chapter. The book is heavily illustrated in black-and-white (which is a pity since a number of the reproduced images are taken from colored originals, and lose much of their visual power). Essentially, Jones has produced a survey of papers from such divergent fields as cognitive psychology, computer graphics, information design, and human-computer interaction, and has prepared them for the optimization community. If we take Jacques Bertin's work as aiming at the geographers' and statisticians' communities, Edward Tufte's two beautiful books as an orientation for many in economics, statistics, and general illustration, and if we grab a good computer graphics textbook, plus one of those marvellous books by graphic designers, we get the same and much more. This reviewer cannot detect more than the explicit targeting on applying visual techniques to optimization that would direct attention to Jones' book. But then it has a number of shortcomings that create doubts: the volume is very high priced (given that there are no colored pictures); it is sloppily produced (with many misspellings, false grammar, poor typersetting, half-empty pages, repetitious bibliographies, unnecessary flip-chart gimmicks); its references are not always first class or most recent, and at times they make the reader wonder why they are included. Above all, its outline is not carried through. Part III contains some material that belongs to part I. The level of discussion changes much: you never really get the essence of what visualization can do to optimization except for adding pictures (what you would have expected in the first place). In summary: although a nice collection of possibilities, I would not recommend it.