Please use this identifier to cite or link to this item:
Title: Analysis of dataflow diagrams by neural network
Authors: Hsu, L.S. 
Chan, S.C. 
Teh, H.H. 
Issue Date: 1988
Citation: Hsu, L.S.,Chan, S.C.,Teh, H.H. (1988). Analysis of dataflow diagrams by neural network. Neural Networks 1 (1 SUPPL) : 447-. ScholarBank@NUS Repository.
Abstract: Given a hardware configuration with a number of processors and a dataflow diagram, one wishes to determine the best way to assign processors to the processes so that the total amount of time needed to complete an algorithm is at a minimum. General algorithm for solving such a problem does not exist. This paper presents a heuristic method which allows use to be made of neural networks to arrange the processes in a definite order. At any time during the execution, processes in the list can be classified into three types of status: 'done', 'ready' and 'waiting'. Whenever a processor has completed its task, it marks the completed process as 'done', check to see if any of the waiting processes can be elevated to 'ready' status, and picks the next ready process in the list for execution. This is repeated until all processes in the list are marked 'done'.
Source Title: Neural Networks
ISSN: 08936080
DOI: 10.1016/0893-6080(88)90471-6
Appears in Collections:Staff Publications

Show full item record
Files in This Item:
There are no files associated with this item.

Page view(s)

checked on Sep 19, 2020

Google ScholarTM



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.