Titles of related interest Simulated Annealing and Boltzmann Machines Emile Aarts, Philips Research Laboratories, Eindhoven, and Eindhoven University of Technology, The Netherlands Jan Korst, Philips Research Laboratories, Eindhoven, The Netherlands Simulated annealing is a solution method in the field of combinatorial optimization based on a simulation of the physical process of annealing. A substantial reduction of the computational effort required to use this method may be achieved by using a computational model based on massively parallel execution, such as the Boltzmann machine, which is a neural network model. This book is intended as an introduction to the theory and applications of simulated annealing and Boltzmann machines. It will be of great interest to students and researchers in combinatorial optimization and neural networks, as well as to all those using optimization techniques in practice. 1988 The Metaphorical Brain 2, Neural Networks and Beyond Michael A. Arbib, Program in Neural, Informational and Behavioral Sciences, University of Southern California, USA ‘This book combines two exciting quests, the quest to understand the workings of the human brain and the quest to build intelligent machines. It shows how each quest can provide insights vital to the success of the other. It develops basic ideas about neural networks, both artificial and biological, and introduces the language of schema theory to describe the distributed interactions that underlie intelligence in the brain of human, animal or robot. It reaffirms the paradigm of highly distributed cooperative computation, showing how it not only deepens our understanding of human mind/brain, but also catalyzes the development of a new generation of computing machinery. The book presents many new results, both from my own group and elsewhere, that have enriched that paradigm during the last fifteen years....The book as a whole, although by no means light reading, should be accessible overall to anyone who reads Scientific American; but it is hoped that much of the material merits the attention not only of "the intelligent laymen" but also of experts and serious students of artificial intelligence, neural networks, robotics, cognitive science, or neuroscience’. —From the Author’s Preface 1989
Les informations fournies dans la section « Synopsis » peuvent faire référence à une autre édition de ce titre.
Focusing on a single neural architecture known as associative memory or Hopfield network, this book explores the different problems that arise in the analysis and design of discrete time and discrete valued recursive networks. In doing so, it provides a simple and structured introduction to this new domain and serves as a guide through the wealth of material scattered among the literature. Topical coverage includes principles, problems and approaches, deterministic and statistical approaches, thermodynamic extension, higher order networks, and network design.From the Inside Flap :
Recursive Neural Networks for Associative Memory Yves Kamp Philips Research Laboratory, Louvain-la-Neuve, Belgium Martin Hasler Ecole Polytechnique Fédérale de Lausanne, Switzerland Neural networks are of increasing interest to a broad sector of the scientific community ranging from neurobiology to electronics and computer science. This book concentrates on a single type of architecture—recursive networks of formal neurons which are used mainly as associative memories (Hopfield, Kohonen). Although simple in structure, they display complex behavioural patterns and there is no one method of explaining all their properties. The authors provide a unifying framework by which different approaches to the subject can be surveyed. Results are currently scattered over different scientific disciplines, each using its own methodology. Emphasis is on theoretical and fundamental issues. An extensive list of references is included. A particularly useful book for postgraduate students and newcomers to this speciality, including professionals and researchers in artificial intelligence, computer science, information processing, electronics, neurobiology and physics.
Les informations fournies dans la section « A propos du livre » peuvent faire référence à une autre édition de ce titre.
Description du livre Wiley, 1990. Hardcover. État : New. book. N° de réf. du libraire 0471928666
Description du livre Wiley, 1990. Hardcover. État : New. book. N° de réf. du libraire 471928666
Description du livre John Wiley & Sons, 1990. Hardcover. État : New. 1. N° de réf. du libraire DADAX0471928666