Please use this identifier to cite or link to this item:
Title: An efficient mapping algorithm of multilayer perceptron on mesh-connected architectures
Authors: Ayoubi, Rafic 
Bayoumi, Magdy A.
Chouemi, A. El
Affiliations: Department of Computer Engineering 
Keywords: Backpropagation
Multilayer perceptrons
Feedforward neural nets
Network topology
Parallel machines
Subjects: Parallel algorithms
Issue Date: 2002
Publisher: IEEE
Part of: Conference Proceedings of the IEEE Fifteenth Anuual International Phoenix Conference on Computers And Communications
Conference: Annual International Phoenix Conference on Computers and Communications (15th : 27-29 March 1996 : Scottsdale, AZ, USA, USA) 
This paper presents a new efficient parallel implementation of neural networks on mesh-connected SRMD machines. A new algorithm to implement the recall and training phases of the multilayer feedforward network with back-error propagation is devised. The developed algorithm is much faster than other known algorithms; it requires O(1) multiplications and O(log N) additions, whereas most others require O(N) multiplications and O(N) additions. The proposed algorithm maximizes parallelism by unfolding the ANN computation to its smallest computational primitives and processes these primitives in parallel.
Ezproxy URL: Link to full text
Type: Conference Paper
Appears in Collections:Department of Computer Engineering

Show full item record

Google ScholarTM


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