Information Technology Reference
In-Depth Information
entities, and neuron pathways as routes (see Fig. 9.1). Within this general informa-
tion processing structure, the major brain areas activated in the transcription typing
task 10 were identified by the following fMRI and PET studies ( [23, 40, 17], see
Fig. 9.1).
Fig. 9.1: The general structure of the queuing network model (QN-MHP) with routes
and servers involved in transcription typing tasks highlighted (server names, brain
structures, and processing logic and time are shown in Table 9.1).
Processing logic and time is based on the literature [10, 27, 38, 14, 37]. If we
consider the network for transcription typing, as shown in Fig. 9.1, upon completing
service at the Pho server, entities have numerous possible routes to follow to traverse
the network: (1) At the Pho server, the entities can choose one of the three routes to
depart the Pho server to the CE, BG, or M1 servers. (2) At the CE server, entities
can choose to move to the BG, SMA, or M1. (3) At the BG server, entities can move
to the SMA or M1 servers. Therefore, there are a total of 3
×
×
=
18 possible
routes for the entities to be processed in the network in transcription typing. An
important question is, therefore, how the entities choose among these routes that
activate (utilize) different brain areas (servers) in different learning stages or when
processing different stimuli at well-learned stages? This question can be answered
by the dynamic part of the model.
3
2
9.2.3.2 The Dynamic Portion of the Queuing Network Model:
Self-Organization of the Queuing Network with Reinforcement
Learning Algorithms
Ungerleider et al. [47] found evidence for the reorganization of brain areas with
practice, which indicates that individual brain areas may change their information
processing speeds in learning. Moreover, some brain areas may have error detection
