Biological Plausibility

Parallel distributed processing (PDP) networks (McClelland & Rumelhart, 1986; Rumelhart & McClelland, 1986) consist of a number of simple processors that perform basic calculations and communicate the results to other processors by sending signals through weighted connections.  The processors operate in parallel, permitting fast computing even when slow componentry is involved.  They exploit implicit, distributed, and redundant representations, making these networks not brittle.  Because networks are not brittle, their behavior degrades gracefully when presented noisy inputs, and such models are damage resistant.  These advantages accrue because artificial neural networks are intentionally biologically plausible.  That is, the basic properties of a PDP network are presumed to reflect functional descriptions of the key elements of how the brain processes information.  That is, connectionists presume that there are extreme differences between how digital computers and brains functions.  However, “these dissimilarities do not imply that brains are not computers, but only that brains are not serial digital computers” (Churchland, Koch, & Sejnowski, 1990, p. 48).  Thus PDP networks are presumed to be biologically plausible because their kind of information processing is more likely to be mapped onto brain processes than is the kind of information processing performed by digital computers.

References:

1. Churchland, P. S., Koch, C., & Sejnowski, T. J. (1990). What is computational neuroscience? In E. L. Schwartz (Ed.), Computational Neuroscience (pp. 46-55). Cambridge, MA: MIT Press.
2. McClelland, J. L., & Rumelhart, D. E. (1986). Parallel Distributed Processing, V.2. Cambridge, MA: MIT Press.
3. Rumelhart, D. E., & McClelland, J. L. (1986). Parallel Distributed Processing, V.1. Cambridge, MA: MIT Press.

(Added April 2011)