No label defined (Q4023489)

This is not a book to learn linear algebra but to apply it. Basic results are recalled in a preliminary chapter. This overview is brief; it contains precise definitions and statements but no proofs. All other chapters are devoted to applications. The main quality of the book lies in the variety of themes: graph theory, Markov chains, linear programming, game theory as well as various economic models (three chapters deal respectively with Leontief input-output matrices, linear systems in economics and von the Neumann input-output model for expanding economies). Every subject is introduced in a clear and concise way; linear algebra appears as an efficient tool to solve practical problems. The concepts and methods are immediately illustrated by simple examples. The book also contains many exercises. The style is appropriately sober. The book has some drawbacks, which seem unavoidable consequences of the qualities pointed out above (namely, variety of themes, clarity and conciseness). First, the applications are presented in elementary and particular setups; few proofs are proposed. The simple, conceptual solutions provided in detail are often only appropriate for problems of ``small size''. General solution methods are seldom described, computation issues are not considered. Furthermore, many applications treated in the book are the very first steps of complex theories, some of which are still growing nowadays (this is specially true for game theory). The author gives few precise references to original work and does not mention recent developments. A crucial point, which may justify the options of the author, is that the book is essentially a textbook for undergraduate students. It does not seem conceivable to teach the material to students who do not know matrix calculus, but the presentation is so clear and self-contained that this could nevertheless make a fairly elementary course. To sum up, the author surveys fundamental, rather classical applications of linear algebra. He gives a good idea of the use of linear methods in economics and social sciences.