Dictionary of Biological Psychology
Connectionism is an approach to developing computer simulations of cognitive processes that attempts to capture the abstract information-processing capabilities of large groups of neurons. Information is represented as distributed patterns of activity over groups of simple, neuron-like processing units.
Co-operative and competitive interactions among units are governed by weighted connections between them; these weights are typically learned using an automatic training procedure (see BACKPROPAGATION; HEBB-LIKE RULE) based on experience with environmental inputs and outputs. Connectionist simulations have successfully modelled phenomena in many perceptual, cognitive, and motor domains.
See also: connectionist neuropsychology; distributed processing; neural networks; parallel processing; serial vs. parallel processing models
Reference
McClelland J.L, Rumelhart D.E., and the PDP Research Group (1986) Parallel Distributed Processing: Explorations in the Microstructure of Cognition, vol. 2, Psychological and Biological Models, MIT Press: Cambridge MA.
DAVID C.PLAUT
This is the complete article, containing 131 words
(approx. 1 page at 300 words per page).
View More Summaries on Connectionism
