Qualitative Analysis and Synthesis of Recursive Artificial Neural Networks

Author: Liu Derong Wang Zhanshan A.N. Michel Zhang Huaguang Ji Ce
Publisher:
Publish Date: 2004-07-01
Features: This book systematically studies the qualitative properties and limitations of recursive artificial neural networks, which are used in the design of (associative memory) and qualitative analysis issues encountered during their implementation. The book consists of nine chapters, with main contents including qualitative analysis of the global and local performance of a series of recursive artificial neural network models, as well as the impact of parameter perturbations, time delays, interconnection structure constraints, and other factors on their performance. The comprehensive design methods for (associative memory) provided in the book include the outer product method, projection learning rule, feature structure method, and perceptron-based training methods, among others. The main feature of this book is its thorough theoretical analysis combined with detailed comprehensive design methods, particularly the innovative comprehensive design methods proposed for recursive artificial neural networks with interconnection structure constraints (including cellular neural networks). This book is suitable for researchers, graduate students, and engineering technicians in applied mathematics, physics, electronic information, automation, and computer application fields who are interested in artificial neural networks.