Abstract
A novel neural network model is described that implements context-dependent learning of complex sequences. The model utilises leaky integrate-and-fire neurons to extract timing information from its input and modifies its weights using a learning rule with synaptic noise. Learning and recall phases are seamlessly integrated so that the network can gradually shift from learning to predicting its input. Experimental results using data from the real-world problem domain demonstrate that the use of context has three important benefits: (a) it prevents catastrophic interference during learning of multiple overlapping sequences, (b) it enables the completion of sequences from missing or noisy patterns, and (c) it provides a mechanism to selectively explore the space of learned sequences during free recall.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
T. Collett K. Fauria K. Dale J. Baron (1997) ArticleTitlePlaces and patterns – a study of context learning in honeybees Journal of Comparative Physiology A 181 343–353
J. Haney K. Lukowiak (2001) ArticleTitleContext learning and the effect of context on memory retrieval in lymnaea Learning and Memory 8 IssueID1 35–43
W. Thach (1996) ArticleTitleOn the specific role of the cerebellum in motor learning and cognition: Clues from pet activation and lesion studies in man Behavioral and Brain Sciences 19 IssueID3 411–431 Occurrence Handle96k:90066
W. Thach H. Goodkin J. Keating (1992) ArticleTitleThe cerebellum and the adaptive coordination of movement Annual Review of Neuroscience 15 403–442 Occurrence Handle10.1146/annurev.ne.15.030192.002155
R. Hirsh (1974) ArticleTitleThe hippocampus and contextual retrieval of information from memory: A theory Behavioural Biology 12 421–444
W. Levy X. Wu (1996) ArticleTitleThe relationship of local context codes to sequence length memory capacity Network: Computation in Neural Systems 7 371–394 Occurrence Handle10.1088/0954-898X/7/2/019
G. Wallenstein H. Eichenbaum M. Hasselmo (1998) ArticleTitleThe hippocampus as an associator of discontiguous events Trends in Neuroscience 21 317–323 Occurrence Handle10.1016/S0166-2236(97)01220-4
M. Chun Y. Jiang (1998) ArticleTitleContextual cueing: Implicit learning and memory of visual context guides spatial attention Cognitive Psychology 36 28–71 Occurrence Handle10.1006/cogp.1998.0681
P. Dominey (1995) ArticleTitleComplex sensory-motor sequence learning based on recurrent state representation and reinforcement learning Biological Cybernetics 73 265–274 Occurrence Handle0828.92003
J. Elman (1990) ArticleTitleFinding structure in time Cognitive Science 14 179–211 Occurrence Handle10.1016/0364-0213(90)90002-E
Shigematsu, Y., Ichikawa, M. and Matsumoto, G.: Reconstitution studies on brain computing with the neural network engineering. In: Perception, Memory and Emotion: Frontiers in Neuroscience, pp. 581–599. Elsevier Science Ltd., 1996.
Milton, J., Mundel, T. and van der Heiden, U.: Traveling activity waves. In: The Handbook of Brain Theory and Neural Network, pp. 994–997. MIT Press, Cambridge, MA., 1995.
Gerstner, W.: Spiking neurons. In: Pulsed Neural Networks, pp. 3–54. MIT Press, 1998.
P. Robinson C. Rennie J. Wright (1997) ArticleTitlePropagation and stability of waves of electrical activity in the cerebral cortex Physical Review E 56 826–841 Occurrence Handle10.1103/PhysRevE.56.826 Occurrence Handle1997PhRvE..56..826R
Mel, B. Why have dendrites? A computational perspective. In: Dendrites, pp. 271–289. Oxford University Press, Oxford, UK, 1999.
Tijsseling A. and Berthouze, L.: A neural network for temporal sequential information. In: Proceedings of the 8th International Conference on Neural Information Processing, Shanghai (China), pp. 1449–1454, 2001.
M. Reiss J. Taylor (1991) ArticleTitleStoring temporal sequences Neural Networks 4 773–787 Occurrence Handle10.1016/0893-6080(91)90057-C
Jordan, M.: Attractor dynamics and parallelism in a connectionist sequential machine. In: Proceedings of the Eight Annual Conference of the Cognitive Science Society, pp. 513–546. Erlbaum: Hillsdale, NJ., 1986.
K. Jim J. Giles B. Horne (1994) Synaptic noise in dynamically-driven recurrent neural networks: Convergence and generalization University of Maryland USA
A. Zador (1998) ArticleTitleImpact of synaptic unreliability on the information transmitted by spiking neurons Journal of Neurophysiology 79 1230–1238
J. Collins C. Chow T. Imhoff (1995) ArticleTitleStochastic resonance without tuning Nature 376 236–238 Occurrence Handle1995Natur.376..236C
French, R.: Dynamically constraining connectionist networks to produce distributed, orthogonal representations to reduce catastrophic interference. In: Proceedings of the 16th Annual Cognitive Science Society Conference, Vol. 5, pp. 207–220, 1994.
I. Tsuda (1992) ArticleTitleDynamic link of memory – chaotic memory map in nonequilibrium neural networks Neural Networks 5 313–326 Occurrence Handle10.1016/S0893-6080(05)80029-2 Occurrence Handle93g:92010
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Berthouze, L., Tijsseling, A. A Neural Model for Context-dependent Sequence Learning. Neural Process Lett 23, 27–45 (2006). https://doi.org/10.1007/s11063-005-2838-x
Issue Date:
DOI: https://doi.org/10.1007/s11063-005-2838-x