Adaptive Analog VLSI Neural Systems

Book excerpt: Adaptive Analog VLSI Neural Systems is the first practical book on neural networks learning chips and systems. It covers the entire process of implementing neural networks in VLSI chips, beginning with the crucial issues of learning algorithms in an analog framework and limited precision effects, and giving actual case studies of working systems. The approach is systems and applications oriented throughout, demonstrating the attractiveness of such an approach for applications such as adaptive pattern recognition and optical character recognition. Dr Jabri and his co-authors from AT&T Bell Laboratories, Bellcore and the University of Sydney provide a comprehensive introduction to VLSI neural networks suitable for research and development staff and advanced students.