New Trends in Neural Computation

Neural computation arises from the capacity of nervous tissue to process information and accumulate knowledge in an intelligent manner. Conventional computational machines have encountered enormous difficulties in duplicatingsuch functionalities. This has given rise to the development of Artificial Neural Networks where computation is distributed over a great number of local processing elements with a high degree of connectivityand in which external programming is replaced with supervised and unsupervised learning. The papers presented in this volume are carefully reviewed versions of the talks delivered at the International Workshop on Artificial Neural Networks (IWANN '93) organized by the Universities of Catalonia and the Spanish Open University at Madrid and held at Barcelona, Spain, in June 1993. The 111 papers are organized in seven sections: biological perspectives, mathematical models, learning, self-organizing networks, neural software, hardware implementation, and applications (in five subsections: signal processing and pattern recognition, communications, artificial vision, control and robotics, and other applications).

Inhalt
Biophysics of neural computation.- Integrated learning in rana computatrix.- A model for centering visual stimuli through adaptive value learning.- A model for the development of neurons selective to visual stimulus size.- An invariant representation mechanism after presynaptic inhibition.- The pancreatic B-cell as a voltage-controlled oscillator.- Approximation of the solution of the dendritic cable equation by a small series of coupled differential equations.- A neural network model inspired in global appreciations about the thalamic reticular nucleus and cerebral cortex connectivity.- Towards more realistic self contained models of neurons: High-order, recurrence and local learning.- McCulloch's neurons revisited.- Biologically motivated approach to face recognition.- Learning by reinforcement: A psychobiological model.- A neural state machine for iconic language representation.- Variable binding using serial order in recurrent neural networks.- Region of influence (ROI) networks. Model and implementation.- A node splitting algorithm that reduces the number of connections in a Hamming distance classifying network.- A high order neural model.- Higher-order networks for the optimization of block designs.- Neural Bayesian classifier.- Constructive methods for a new classifier based on a radial-basis-function neural network accelerated by a tree.- Practical realization of a radial basis function network for handwritten digit recognition.- Design of fully and partially connected random neural networks for pattern completion.- Representation and recognition of regular grammars by means of second-order recurrent neural networks.- Connectionist models for syllabic recognition in the time domain.- Sparsely interconnected artificial neural networks for associative memories.- Dynamic analysis of networks of neural oscillators.- Optimised attractor neural networks with external inputs.- Non-orthogonal bases and metric tensors: An application to artificial neural networks.- Genetic synthesis of discrete-time recurrent neural network.- Optimization of a competitive learning neural network by genetic algorithms.- Adaptive models in neural networks.- Self-organizing grammar induction using a neural network model.- The role of forgetting in efficient learning strategies for self-organising discriminator-based systems.- Simulation of stochastic regular grammars through simple recurrent networks.- Local stochastic competition and vector quantization.- MHC An evolutive connectionist model for hybrid training.- Fast-convergence learning algorithms for multi-level and binary neurons and solution of some image processing problems.- Invariant object recognition using fahlman and Lebiere's learning algorithm.- Realization of subjective correspondence in artificial neural network trained by Fahlman and Lebiere's learning algorithm.- Bimodal distribution removal.- A simplified ARTMAP architecture for real-time learning.- B-Learning: A reinforcement learning algorithm, comparison with dynamic programming.- Increased complexity training.- Optimized learning for improving the evolution of piecewise linear separation incremental algorithms.- A method of pruning layered feed-forward neural networks.- Tests of different regularization terms in small networks.- On the distribution of feature space in Self-Organising mapping and convergence accelerating by a Kalman algorithm.- A learning algorithm to obtain self-organizing maps using fixed neighbourhood Kohonen networks.- Analysing a contingency table with Kohonen maps: A factorial correspondence analysis.-Dynamics of self-organized feature mapping.- Comparative study of self-organizing neural networks.- GANNet: A genetic algorithm for optimizing topology and weights in neural network design.- Vector quantization and projection neural network.- Constructive design of LVQ and DSM classifiers.- Linear vector classification: An improvement on LVQ algorithms to create classes of patterns.- Non-greedy adaptive vector quantizers.- Hybrid programming environments.- Automatic generation of C++ code for neural network simulation.- Urano: An object-oriented artificial neural network simulation tool.- Realistic simulation tool for early visual processing including space, time and colour data.- Language supported storage and reuse of persistent neural network objects.- Flexible operating environment for matrix based neurocomputers.- A parallel implementation of kohonen's self-organizing maps on the smart neurocomputer.- Simulation of neural networks in a distributed computing environment using Neurograph.- Full automatic ann design: A genetic approach.- Hardware implementations of artificial neural networks.- A neural network chip using CPWM modulation.- Hardware implementation of a neural network for high energy physics application.- MapA: An array processor architecture for neural networks.- Limitation of connectionism in MLP.- High level synthesis of neural network chips.- Neural network simulations on massively parallel computers: Applications in chemical physics.- A model based approach to the performance analysis of multi-layer networks realised in linear systolic arrays.- The temporal noisy-leaky integrator neuron with additional inhibitory inputs.- Architectures for self-learning neural network modules.- The generic neuron architectural framework for the automatic generationof ASICs.- A risc architecture to support neural net simulation.- Hardware design for self organizing feature maps with binary input vectors.- The Kolmogorov signal processor.- Projectivity invariant pattern recognition with high-order neural networks.- Rejection of incorrect answers from a neural net classifier.- Nonlinear time series modeling by competitive segmentation of state space.- Identification and prediction of non-linear models with recurrent neural network.- Use of unsupervised neural networks for classification of blood pressure time series.- Application of artificial neural networks to chest image classification.- Combination of self-organizing maps and multilayer perceptrons for speaker independent isolated word recognition.- An industrial application of neural networks to natural textures classification.- Use of a layered neural nets as a display method for n-dimensional distributions.- MLP modular versus yprel classifiers.- How many hidden neurons are needed to recognize a symmetrical pattern ?.- Hopfield neural network for routing.- Neural network routing controller for communication parallel multistage interconnection networks.- Adaptive routing using cellular automata.- Optimal blind equalization of Gaussian channels.- Noise prediction in urban traffic by a neural approach.- A connectionist approach to the correspondence problem in computer vision.- Self-organizing feature m…