Parallel
Processing: The Connectionist Model
Computer-inspired
information-processing theories assume that humans, like computers, process
information serially. That is, information is processed one step after another.
These aspects involve parallel processing, in which multiple operations go on all at once. Therefore, the
distribution of parallel processes better explains the speed and accuracy of
human information processing.
As
a result of these considerations, many contemporary models of knowledge
representation emphasize the importance of parallel processing in human
cognition. As a further result of
interest in parallel processing, some computers have been made to simulate
parallel processing, such as through so-called neural networks of
interlinked
computer processors.
At
present, many cognitive psychologists are exploring the limits of parallel
processing models. According to parallel
distributed processing (PDP) models or
connectionist models, we handle very large numbers of cognitive operations at once
through a network distributed across incalculable numbers of locations in the
brain.
How the PDP Model Works
The
mental structure within which parallel processing is believed to occur is a
network. In connectionist networks, all forms of knowledge are represented
within the network structure. Recall that the fundamental element of the
network is the node. Each node is connected to many other nodes. These
interconnected patterns of nodes enable the individual to organize meaningfully
the knowledge contained in the connections among the various nodes. In many
network models, each node represents a concept.
The
network of the PDP model is different in key respects from the semantic
network. In the PDP model, the network
comprises neuron-like units. They do
not, in and of themselves, actually represent concepts, propositions, or
any
other type of information. Thus, the pattern of connections represents the
knowledge, not the specific units.
Parallel Distributed Processing Models (PDP)
A class of neurally
inspired information processing models that attempt to model information
processing the way it actually takes place in the brain.
This model was
developed because of findings that a system of neural connections appeared to
be distributed in a parallel array in addition to serial pathways. As such,
different types of mental processing are considered to be distributed
throughout a highly complex neuronetwork.
The PDP model has 3
basic principles: a.) the representation of information is distributed (not local)
b.) memory and knowledge for specific things are not stored explicitly, but
stored in the connections between units. c.) learning can occur with gradual
changes in connection strength by experience.
"These models
assume that information processing takes place through interactions of large
numbers of simple processing elementscalled units, each sending excitatory and
inhibitory signals to other units." (McLelland, J., Rumelhart, D., &
Hinton, G., 1986,p.10)
Rumelhart, Hinton,
and McClelland (1986) state that there are 8 major components of the PDP model
framework:
1.) a set of
processing units
2.) a state of
activation
3.) an output
function for each unit
4.) a pattern of
connectivity among units
5.) a propagation
rule for propagating patterns of activities through the network of
connectivities
6.) an activation
rule for combining the inputs impinging on a unit with the current state of
that unit to produce a new level of activation for the unit
7.) a learning rule
whereby patterns of connectivity are modified by experience
8.) an environment
within which the system must operate
No comments:
Post a Comment