Abstract
A novel artificial neural network, derived from neurobiological observations, is described and examples of its performance are presented. This DYnamically STable Associative Learning (DYSTAL) network associatively learns both correlations and anticorrelations, and can be configured to classify or restore patterns with only a change in the number of output units. DYSTAL exhibits some particularly desirable properties: computational effort scales linearly with the number of connections, i.e., it is0(N) in complexity; performance of the network is stable with respect to network parameters over wide ranges of their values and over the size of the input field; storage of a very large number of patterns is possible; patterns need not be orthogonal; network connections are not restricted to multi-layer feed-forward or any other specific structure; and, for a known set of deterministic input patterns, the network weights can be computed, a priori, in closed form. The network has been associatively trained to perform the XOR function as well as other classification tasks. The network has also been trained to restore patterns obscured by binary or analog noise. Neither global nor local feedback connections are required during learning; hence the network is particularly suitable for hardware (VLSI) implementation.
Similar content being viewed by others
References
Akaike H (1959) On a successive transformation of probability distribution and its application to the analysis of the optimal gradient method. Ann Inst Stat Math Tokyo 11:1–16
Alkon DL (1983) Learning in a marine snail. Sci Am 249:70–84
Alkon DL (1984) Calcium-mediated reduction of ionic currents: a biophysical memory trace. Science 226:1037–1045
Alkon DL (1987) Memory traces in the brain. Cambridge University Press, Cambridge
Alkon DL (1989) Memory storage and neural systems. Sci Am July: 42–50
Alkon DL, Rasmussen H (1988) A spatial-temporal model of cell activation. Science 239:998–1005
Alkon DL, Quek F, Vogl TP (1988) Computer modeling of associative learning. In: Touretzky DS (ed) Advances in neural information processing systems I. Morgan-Kaufmann, San Mateo, Calif
Anderson JA (1983) Cognitive and psychological computation with neural models. IEEE Trans SMC-13:799–815
Anderson JA (1986) Cognitive capabilities of a parallel system. In: Bienenstock E, Fegelman Soulie F, Weisbuch G (eds) Disordered systems and biological organization. Springer, Berlin Heidelberg New York
Bank B, Gurd JW, Chute DL (1986) Decreased phosphorylation of synaptic glycoproteins following hippocampal kindling. Brain Res 399:390–394
Bank B, DeWeer A, Kuzirian AM, Rasmussen H, Alkon DL (1988) Classical conditioning induces long-term translocation of protein kinase C in rabbit hippocampal CA1 cells. Proc Natl Acad Sci USA 85:1988–1992
Chiang T, Chow Y (1988) On eigenvalues and annealing rates. Math Op Res 13:508–511
Coulter DA, Lo Turco JJ, Kubota M, Disterhoft JF, Alkon DL (1989) Classical conditioning reduces amplitude and duration of calcium-dependent after hyperpolarization in rabbit hippocampal pyramidal cells. J Neurophysiol 61:971–981
Dåhlquist G, Björck A (1974) Numerical methods. Prentice-Hall, Englewood Cliffs
Disterhoft JF, Coulter DA, Alkon DL (1986) Conditioning-specific membrane changes of rabbit hippocampal neurons measured in vitro. Proc Natl Acad Sci USA 83:2733–2737
Fukushima K (1988) Neocognitron: A Hierarchical Neural Network Capable of Visual Pattern Recognition. Neural Netw 1:119–130
Grossberg S (1982) Studies of mind and brain. Reidel, Dordrecht
Grossberg S (1987) Competitive learning: from interactive activation to adaptive resonance. Cogn Sci 11:23–63
Hebb DO (1949) The organization of behavior. Wiley, New York
Hopfield JJ (1982) Neural networks and physical systems with emergent collective computational abilities. Proc Natl Acad Sci USA 79:2554–2558
Hopfield JJ, Tank DW (1986) Computing with neural circuits: a model. Science 233:625–633
Kienker PK, Sejnowski TJ, Hinton GE, Schumacher LE (1986) Separating figure from ground with parallel network. Perception 15:197–216
Klopf H (1988) A neuronal model of classical conditioning. Psychobiology 16:85–125
Kohonen T, Makisara K (1989) The self-organizing feature maps. Phys Scr 39:168–172
LoTurco JL, Coulter DA, Alkon DL (1988) Enhancement of synaptic potentials in rabbit CA1 pyramidal neurons following classical conditioning. Proc Natl Acad Sci USA 85:1672–1676
McClelland JL (1985) Putting knowledge in its place: a scheme for programming parallel processing structures on the fly. Cogn Sci 9:113–146
McClelland JL, Rumelhart DE (1981) An interactive activation model of context effects in letter perception: part 1. An account of basic findings. Psychol Rev 88:375–407
Olds JL, Anderson ML, McPhie DL, Staten LD, Alkon DL (1989) Imaging memory-specific changes in the distribution of protein kinase C within the Hippocampus. Science 245:866–869
Press WH, Flannery BP, Teukolsky SA, Vetterling WT (1986) Numerical recipes, Cambridge University Press, Cambridge
Rosenblatt F (1959) Two theorems of statistical separability in the perceptron. In: Mechanization of thought processes: Proceedings of a symposium held at the National Physical Laboratory, November 1958. 1:421–456 HM Stationary Office
Roth MW (1988) Neural network technology and its applications. Johns Hopkins APL Tech Digest 9:242–253
Rumelhart DE, McClelland JL (1986) (eds) Parallel distributed processing: explorations in the microstructure of cognition, vol I: Foundations; vol II: Applications. MIT Press, Cambridge
Rumelhart DE, Zipser D (1985) Feature discovery by competitive learning. Cogn Sci 9:75–112
Sejnowski TJ, Kienker PK (1986) Learning symmetry groups with hidden units: Beyond the perceptron. Physica D 22:260–275
Sejnowski TJ, Rosenberg CR (1986) NETtalk: A parallel network that learns to read aloud. The Johns-Hopkins University Electrical Engineering and Computer Technical Report JHU/EECS-86/01
Vogl TP, Mangis JK, Rigler AK, Zink WT, Alkon DL (1988) Accelerating the convergence of the back-propagation method. Biol Cybern 59:257–263
Zink WT, Vogl TP, Mangis JK (1988) Neural networks as classifiers of noisy patterns: an experimental comparison with Bayesian classifiers. Presented at the First National Conference of the International Neural Network Society, Boston, Mass
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Alkon, D.L., Blackwell, K.T., Barbour, G.S. et al. Pattern-recognition by an artificial network derived from biologic neuronal systems. Biol. Cybern. 62, 363–376 (1990). https://doi.org/10.1007/BF00197642
Received:
Accepted:
Issue Date:
DOI: https://doi.org/10.1007/BF00197642