A massively parallel architecture for a self-organizing neural pattern recognition machine
Computer Vision, Graphics, and Image Processing
Learning internal representations by error propagation
Parallel distributed processing: explorations in the microstructure of cognition, vol. 1
Journal of Cognitive Neuroscience
Hi-index | 0.00 |
An avalanche model of motor sequence encoding is presented. The ultimate aim is to reproduce data showing that monkeys with frontal lobe damage can learn an invariant sequence of movements if it is rewarded, but cannot learn to perform any one of several variations of a sequence if all are rewarded. In this article, we present simulations of the primary learning of sequences, and demonstrate parameters that can lead to a primacy effect in recalling items of this sequence from long-term memory. Two different versions of the avalanche network and their simulations are presented, of which the second version includes a layer of sequence detectors. Suggestions are made for including yet another layer to classify these temporal sequences and group them together based on reward. Analogies are drawn between the classifier layer and the frontal lobes, and between the avalanche module and part of the basal ganglia.