Complex valued neural networks

Más información sobre el libro

This second enlarged and revised edition of the successful monograph on complex-valued neural networks (CVNNs) is suitable for graduate and undergraduate courses across various fields, including electrical engineering, informatics, and robotics. The updated edition incorporates recent trends in CVNN research, nearly doubling the number of references. It discusses the parametron, invented in 1954, and contrasts the advantages of CVNNs with real-valued neural networks. The book serves as a valuable resource for newcomers studying adaptive signal processing, robotics inspired by human neural systems, and brain-like information processing, as well as for interdisciplinary research aimed at creating a comfortable society. It also supports professionals engaged in research and development of new products and services. The author aims to provide readers with practical insights into leveraging neural networks effectively. Key questions addressed include the necessity of considering neural networks that utilize complex numbers, the specific advantages of CVNNs, their origins, and their principal applications. Through detailed descriptions and examples, the book seeks to inspire readers with innovative ideas and a deeper understanding of the subject.