In particular we examine what happens when the ntunber of layers is large or when the connectivity between layers is local and investigate some of the properties of an autoassociative algorithm. Notation will be as in [1] where additional motivations and references can be found. It is usual to criticize linear networks because "linear functions do not compute" and because several layers can always be reduced to one by the proper multiplication of matrices. However this is not the point of view adopted here. It is assumed that the architecture of the network is given (and could perhaps depend on external constraints) and the purpose is to understand what happens during the learning phase, what strategies are adopted by a synaptic weights modifying algorithm, ... [see also Cottrell et al. (1988) for an example of an application andthe work of Linsker (1988) on the emergence of feature detecting units in linear networks}.

The application of neural networks to the demodulation of spread-spectrum signals in a multiple-access environment is considered. This study is motivated in large part by the fact that, in a multiuser system, the conventional (matched filter) receiversuffers severe performance degradation as the relative powers of the interfering signals become large (the "near-far" problem). Furthermore, the optimum receiver, which alleviates the near-far problem, is too complex to be of practical use. Receivers based on multi-layer perceptrons are considered as a simple and robust alternative to the optimum solution.The optimum receiver is used to benchmark the performance of the neural net receiver; in particular, it is proven to be instrumental in identifying the decision regions of the neural networks. The back-propagation algorithm and a modified version of it are used to train the neural net. An importance sampling technique is introduced to reduce the number of simulations necessary to evaluate the performance of neural nets.

Linsker's model was presented with input patterns each consisting

Learning procedures that measure how random perturbations of unit activities correlatewith changes in reinforcement are inefficient but simple to implement in hardware. Procedures like back-propagation (Rumelhart, Hinton and Williams, 1986) which compute how changes in activities affect theoutput error are much more efficient, but require more complex hardware. GEMINI is a hybrid procedure for multilayer networks, which shares many of the implementation advantages of correlational reinforcement proceduresbut is more efficient. GEMINI injects noise only at the first hidden layer and measures the resultant effect on the output error. A linear network associated with each hidden layer iteratively inverts the matrix which relates the noise to the error change, thereby obtaining the error-derivatives. No back-propagation is involved, thus allowing unknown non-linearitiesin the system. Two simulations demonstrate the effectiveness of GEMINI.

This paper describes an analog version of a self-organizing feature map circuit. The design implements Kohonen's self-organizing feature map algorithm [Kohonen, 1988] with some modifications imposed by practical circuit limitations. The feature map algorithm automatically adapts connection weights to nodes in the network such that each node comes to represent a distinct class of features in the input space. The system also self-organizes such that neighboring nodes become responsive to similar input classes. The prototype circuit was fabricated in two parts (for testability); a 4 node, 4 input synaptic array, and a weight adaptation and refresh circuit.

Bill Baird Department of Biophysics U.C. Berkeley ABSTRACT A new learning algorithm for the storage of static and periodic attractors in biologically inspired recurrent analog neural networks is introduced. For a network of n nodes, n static or n/2 periodic attractors may be stored. The algorithm allows programming of the network vector field independent ofthe patterns to be stored. Stability of patterns, basin geometry, and rates of convergence may be controlled. Standing or traveling wave cycles may be stored to mimic the kind of oscillating spatial patterns that appear in the neural activity of the olfactory bulb and prepyriform cortex during inspiration and suffice, in the bulb, to predict the pattern recognition behavior of rabbits in classical conditioning experiments.

CA 94305 ABSTRACT The potential of adaptive networks to learn categorization rules and to model human performance is studied by comparing how natural and artificial systems respond to new inputs, i.e., how they generalize. Like humans, networks can learn a detenninistic categorization task by a variety of alternative individual solutions. An analysis of the constraints imposedby using networks with the minimal number of hidden units shows that this "minimal configuration" constraint is not sufficient A further analysis of human and network generalizations indicates that initial conditions may provide important constraints on generalization. A new technique, which we call "reversed learning", is described for finding appropriate initial conditions. INTRODUCTION We are investigating the potential of adaptive networks to learn categorization tasks and to model human performance.

We have fabricated a test chip in 2 micron CMOS technology that embodies these ideas and we report our evaluation of the microchip and our plans for improvements. Knowledge is encoded in the test chip by presenting digital patterns to it that are examples of a desired input-output Boolean mapping. This knowledge is learned and stored entirely on chip in a digitally controlled synapse-like element in the form of connection strengths between neuron-like elements. The only portion of this learning system which is off chip is the VLSI test equipment used to present the patterns. This learning system uses a modified Boltzmann machine algorithm[3] which, if simulated on a serial digital computer, takes enormous amounts of computer time. Our physical implementation is about 100,000 times faster. The test chip, if expanded to a board-level system of thousands of neurons, would be an appropriate architecture for solving artificial intelligence problems whose solutions are hard to specify using a conventional rule-based approach. Examples include speech and pattern recognition and encoding some types of expert knowledge.

This paper proposes a means of using the knowledge in a network to determine the functionality or relevance of individual units, both for the purpose of understanding the network's behavior and improving its performance. The basic idea is to iteratively train the network to a certain performancecriterion, compute a measure of relevance that identifies whichinput or hidden units are most critical to performance, and automatically trim the least relevant units. This skeletonization technique canbe used to simplify networks by eliminating units that convey redundant information; to improve learning performance by first learning with spare hidden units and then trimming the unnecessary ones away, thereby constraining generalization; and to understand the behavior of networks in terms of minimal "rules."

I have previously proposed [Linsker, 1987, 1988] a principle of "maximum information preservation," also called the "infomax" principle, that may account for certain aspects of the organization of a layered perceptual network. The principle applies to a layer L of cells (which may be the input layer or an intermediate layer of the network) that provides input to a next layer M. The mapping of the input signal vector L onto an output signal vector M, f:L M, is characterized by a conditional probability density function ("pdf") p(MI L).