This volume contains selected papers from WIRN VIETRI-96, the 8th Italian Workshop on Neural Nets, held Vietri sul Mare, Salerno, Italy, from 23-25 May 1996. The papers cover a variety of topics related to neural networks, including pattern recognition, signal processing, theoretical models, applications in science and industry, virtual reality, fuzzy systems, and software algorithms. By providing the reader with a comprehensive overview of recent research work in this area, the volume makes an invaluable contribution to the Perspectives in Neural Computing Series. Neural Nets - WIRN VIETRI-96 will provide invaluable reading material for anyone who needs to keep up to date with the latest developments in neural networks and related areas. It will be of particular interest to academic and industrial researchers, and postgraduate and graduate students.
Table of ContentsSection 1 - Invited Paper.- A Class of Cost Functions for Independence.- Section 2 - Review Papers.- Virtual Reality and Neural Networks.- Recent Results in On-Line Prediction and Boosting.- Section 3 - Eduardo R. Caianiello Lectures.- Input/Output HMMs: A Recurrent Bayesian Network View.- Neural Nonlinear Controllers and Observers:Stability Results.- Section 4 - Mathematical Models.- Neural Network Fuzzification: A Critical Review of the Fuzzy Learning Vector Quantization Model.- Cultural Evolution in a Population of Neural Networks.- A New Incremental Learning Technique.- Solving Algebraic and Geometrical Problems Using Neural Networks.- A Bayesian Framework for Associative Memories.- Proof of the Universal Approximation of a Set of Fuzzy Functions.- The Computational Neural Map and its Capacity.- Classification of Structures by Recurrent Neural Networks.- Section 5 - Pattern Recognition and Robotics.- Neuro-Fuzzy Processing of Remote Sensed Data.- The Use of Neural Networks for the Automatic Detection and Classification of Weak Photometric Sub-Components in Early-Type Galaxies.- A Generalized Regularization Network for Remote Sensing Data Classification.- A Hybrid Neural Network Architecture for Dynamic Scene Understanding.- Proposal of a Darwin-Neural Network for a Robot Implementation.- A Novel Hypothesis on Cortical Map: Topological Continuity.- A Mlp-Based Digit and Uppercase Characters Recognition System.- Are Multilayer Perceptrons Adequate for Pattern Recognition and Verification?.- Computational Intelligence in Electromagnetics.- M.A.I.A. Unsupervisioned Neural Network in the Compression and Decompression of Images.- Fast Spline Neural Networks for Image Compression.- Age Estimates of Stellar Systems by Artificial Neural Networks.- Section 6 - Applications.- Neural Nets for Hybrid On-Line Plant Control.- Using Fuzzy Logic to Solve Optimization Problems by Hopfield Neural Model.- FIR NNs and Time Series Prediction: Applications to Stock Market Forecasting.- Simulation of Traffic Flows in Transportation Nets with Unsupervisioned MAIA Neural Net.- Towards Fog Forecasting in Meteorology by Means of Back-Propagation Neural Networks with Weighted Least-Squares Training.- Spectral Mapping: A Comparison of Connectionist Approaches.- Section 7 - Architectures and Algorithms.- A Reconfigurable Analog VLSI Neural Network Architecture with Non Linear Synapses.- Off-Chip Training of Analog Hardware Feed-Forward Neural Networks through Hyper - Floating Resilient Propagation.- A Learning Strategy which Increases Partial Fault Tolerance of Neural Nets.- An Integrated Neural and Algorithmic System for Optical Flow Computation.- FIR NNs and Temporal BP: Implementation on the Meiko CS-2.- Some Comments and Experimental Results on Bayesian Regularization.- Constructive Fuzzy Neural Networks.- An Adaptable Boolean Neural Net Trainable to Comment on its Own Innerworkings.- Fast Training of Recurrent Neural Networks by the Recursive Least Squares Method.- Author Index.