Department of Electrical and Computer Engineering University of Notre Dame Notre Dame, IN 46556 ABSTRACT In the present paper we survey and utilize results from the qualitative theory of large scale interconnected dynamical systems in order to develop a qualitative theory for the Hopfield model of neural networks. In our approach we view such networks as an interconnection of many single neurons. Our results are phrased in terms of the qualitative properties of the individual neurons and in terms of the properties of the interconnecting structure of the neural networks. Aspects of neural networks which we address include asymptotic stability, exponential stability, and instability of an equilibrium; estimates of trajectory bounds; estimates of the domain of attraction of an asymptotically stable equilibrium; and stability of neural networks under structural perturbations. INTRODUCTION In recent years, neural networks have attracted considerable attention as candidates for novel computational systemsl-3.

Current Focus Of Learning Research Most connectionist learning algorithms may be grouped into three general catagories, commonly referred to as supenJised, unsupenJised, and reinforcement learning. Supervised learning requires the explicit participation of an intelligent teacher, usually to provide the learning system with task-relevant input-output pairs (for two recent examples, see [1,2]). Unsupervised learning, exemplified by "clustering" algorithms, are generally concerned with detecting structure in a stream of input patterns [3,4,5,6,7]. In its final state, an unsupervised learning system will typically represent the discovered structure as a set of categories representing regions of the input space, or, more generally, as a mapping from the input space into a space of lower dimension that is somehow better suited to the task at hand. In reinforcement learning, a "critic" rewards or penalizes the learning system, until the system ultimately produces the correct output in response to a given input pattern [8]. It has seemed an inevitable tradeoff that systems needing to rapidly learn specific, behaviorally useful input-output mappings must necessarily do so under the auspices of an intelligent teacher with a ready supply of task-relevant training examples. This state of affairs has seemed somewhat paradoxical, since the processes of Rerceptual and cognitive development in human infants, for example, do not depend on the moment by moment intervention of a teacher of any sort. Learning by Doing The current work has been focused on a fourth type of learning algorithm, i.e. learning-bydoing, an approach that has been very little studied from either a connectionist perspective

Instead, a signal space approach is used to gain new insights via ease of analysis and geometrical interpretation.

In a useful associative memory, an initial state should lead reliably to the "closest" memory. This requirement suggests that a well-behaved basin of attraction should evenly surround its attractor and have a smooth and regular shape. One dimensional basin maps plotting "pull in" probability against Hamming distance from an attract or do not reveal the shape of the basin in the high dimensional space of initial states9.

It is clear that conventional computers lag far behind organic computers when it comes to dealing with very large data rates in problems such as computer vision and speech recognition.

The nets are also described as parallel distributed systems!, connectionist models

REFLEXIVE ASSOCIATIVE MEMORIES Hendrlcus G. Loos Laguna Research Laboratory, Fallbrook, CA 92028-9765 ABSTRACT In the synchronous discrete model, the average memory capacity of bidirectional associative memories (BAMs) is compared with that of Hopfield memories, by means of a calculat10n of the percentage of good recall for 100 random BAMs of dimension 64x64, for different numbers of stored vectors. The memory capac1ty Is found to be much smal1er than the Kosko upper bound, which Is the lesser of the two dimensions of the BAM. On the average, a 64x64 BAM has about 68 % of the capacity of the corresponding Hopfield memory with the same number of neurons. The memory capacity limitations are due to spurious stable states, which arise In BAMs In much the same way as in Hopfleld memories. Occurrence of spurious stable states can be avoided by replacing the thresholding in the backlayer of the BAM by another nonl1near process, here called "Dominant Label Selection" (DLS).

DISTRIBUTED NEURAL INFORMATION PROCESSING IN THE VESTIBULO-OCULAR SYSTEM Clifford Lau Office of Naval Research Detach ment Pasadena, CA 91106 Vicente Honrubia* UCLA Division of Head and Neck Surgery Los Angeles, CA 90024 ABSTRACT A new distributed neural information-processing model is proposed to explain the response characteristics of the vestibulo-ocular system and to reflect more accurately the latest anatomical and neurophysiological data on the vestibular afferent fibers and vestibular nuclei. In this model, head motion is sensed topographically by hair cells in the semicircular canals. Hair cell signals are then processed by multiple synapses in the primary afferent neurons which exhibit a continuum of varying dynamics. The model is an application of the concept of "multilayered" neural networks to the description of findings in the bullfrog vestibular nerve, and allows us to formulate mathematically the behavior of an assembly of neurons whose physiological characteristics vary according to their anatomical properties. INTRODUCTION Traditionally the physiological properties of individual vestibular afferent neurons have been modeled as a linear time-invariant system based on Steinhausents description of cupular motion.

How Neural Nets Work Alan Lapedes Robert Farber Theoretical Division Los Alamos National Laboratory Los Alamos, NM 87545 Abstract: There is presently great interest in the abilities of neural networks to mimic "qualitative reasoning" by manipulating neural incodings of symbols. Less work has been performed on using neural networks to process floating point numbers and it is sometimes stated that neural networks are somehow inherently inaccurate and therefore best suited for "fuzzy" qualitative reasoning. Nevertheless, the potential speed of massively parallel operations make neural net "number crunching" an interesting topic to explore. In this paper we discuss some of our work in which we demonstrate that for certain applications neural networks can achieve significantly higher numerical accuracy than more conventional techniques. In particular, prediction of future values of a chaotic time series can be performed with exceptionally high accuracy. We analyze how a neural net is able to do this, and in the process show that a large class of functions from Rn. Rffl may be accurately approximated by a backpropagation neural net with just two "hidden" layers. The network uses this functional approximation to perform either interpolation (signal processing applications) or extrapolation (symbol processing applicationsJ.

To illustrate the extent of this additional complexity, Fig. Input 6a shows a schematized circuit for a computational unit of method V