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).

A pool of handwritten signatures is used to train a neural network forthe task of deciding whether or not a given signature is a forgery. The network is a feedforward net, with a binary image as input. There is a hidden layer, with a single unit output layer. The weights are adjusted according to the backpropagation algorithm. The signatures are entered into a C software program through the use of a Datacopy Electronic Digitizing Camera. The binary signatures arenormalized and centered. The performance is examined as a function of the training set and network structure. The best scores are on the order of 2% true signature rejection with 2-4% false signature acceptance.

ABSTRACT This report describes the design of a programmable general purpose analog neural computer and simulator. It is intended primarily for real-world real-time computations such as analysis of visual or acoustical patterns, robotics and the development of special purpose neural nets. The machine is scalable and composed of interconnected modules containing arrays of neurons, modifiable synapses and switches. It runs entirely in analog mode but connection architecture, synaptic gains and time constants as well as neuron parameters are set digitally. Each neuron has a limited number of inputs and can be connected to any but not all other neurons.

We propose a new neural network model and its learning algorithm. The proposed neural network consists of four layers - input, hidden, output and final output layers. The hidden and output layers are multiple. Using the proposed SICL(Spread Pattern Information and Cooperative Learning) algorithm, it is possible to learn analog data accurately and to obtain smooth outputs. Using this neural network, we have developed a speech production system consisting of a phonemic symbol production subsystem and a speech parameter production subsystem. We have succeeded in producing natural speech waves with high accuracy.

This paper describes an ongoing effort which approaches neural net research in a program of close collaboration of neurosc i ent i sts and eng i neers. The effort is des i gned to elucidate associative learning in the marine snail Hermissenda crassicornist in which Pavlovian conditioning has been observed. Learning has been isolated in the four neuron network at the convergence of the v i sua 1 and vestibular pathways in this animal t and biophysical changes t specific to learning t have been observed in the membrane of the photoreceptor B cell. A basic charging capacitance model of a neuron is used and enhanced with biologically plausible mechanisms that are necessary to replicate the effect of learning at the cellular level. These mechanisms are nonlinear and are t primarilYt instances of second order control systems (e.g.

We propose a new neural network structure that is compatible with silicon technology and has built-in learning capability. The thrust of this network work is a new synapse function. The synapses have the feature that the learning parameter is embodied inthe thresholds of MOSFET devices and is local in character. Thenetwork is shown to be capable of learning by example as well as exhibiting the desirable features of the Hopfield type networks. The thrust of what we want to discuss is a new synapse function for an artificial neuron to be used in a neural network.

Details of a number of image recognition tasks, including character recognition and object centring, are described.

Learning procedures that measure how random perturbations of unit activities correlate with 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 the output 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 procedures but 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-linearities in the system. Two simulations demonstrate the effectiveness of GEMINI.

An improved learning paradigm that offers a significant reduction in computation time during the supervised learning phase is described. It is based on extending the role that the neuron plays in artificial neural systems. Prior work has regarded the neuron as a strictly passive, nonlinear processing element, and the synapse on the other hand as the primary source of information processing and knowledge retention. In this work, the role of the neuron is extended insofar as allowing its parameters to adaptively participate in the learning phase. The temperature of the sigmoid function is an example of such a parameter.