DocumentCode
329110
Title
Efficient neural network training algorithm for the Cray Y-MP supercomputer
Author
Leung, Chung Siu ; Setiono, Rudy
Author_Institution
Dept. of Inf. Syst. & Comput. Sci., Nat. Univ. of Singapore, Singapore
Volume
2
fYear
1993
fDate
25-29 Oct. 1993
Firstpage
1943
Abstract
An efficient implementation of a quasi-Newton algorithm for feedforward neural network training on a Cray Y-MP is presented. The most time-consuming step of a neural network training using the quasi-Newton algorithm is the computation of the error function and its gradient. We describe in this paper how this step can be implemented so that the neural network training may take full advantage of the Cray vectorization capabilities.
Keywords
Cray computers; Newton method; feedforward neural nets; learning (artificial intelligence); parallel algorithms; parallel processing; Cray Y-MP supercomputer; error function; feedforward neural network; gradient calculation; neural network training; quasi-Newton algorithm; vectorization; Computer networks; Computer science; Constraint optimization; Feedforward neural networks; Information systems; Minimization methods; Network topology; Neural networks; Optimization methods; Supercomputers;
fLanguage
English
Publisher
ieee
Conference_Titel
Neural Networks, 1993. IJCNN '93-Nagoya. Proceedings of 1993 International Joint Conference on
Print_ISBN
0-7803-1421-2
Type
conf
DOI
10.1109/IJCNN.1993.717036
Filename
717036
Link To Document