Title :
Improving performances of Battiti-Shanno´s quasi-Newtonian algorithms for learning in feed-forward neural networks
Author :
Fanelli, S. ; Paparo, P. ; Protasi, M.
Author_Institution :
Dipartimento di Matematica, Rome Univ., Italy
fDate :
29 Nov-2 Dec 1994
Abstract :
The authors describe a new improved Quasi-Newtonian algorithm (named OSSV) for effective learning in MLP-networks. OSSV, which is a variant of Battiti-Shanno´s original OSS method, is able to speed-up the convergence process of the network, maintaining an O(N) complexity. Numerical results show that by OSSV the computational effort of the original OSS can be reduced by a factor increasing with the number of epochs
Keywords :
computational complexity; convergence of numerical methods; feedforward neural nets; learning (artificial intelligence); multilayer perceptrons; MLP-networks; OSSV; complexity; computational effort; convergence process; feedforward neural networks; learning; multilayer perceptron; quasi-Newtonian algorithms; Acceleration; Convergence; Feedforward neural networks; Feedforward systems; Gradient methods; Intelligent networks; Iterative algorithms; Iterative methods; Minimization methods; Neural networks;
Conference_Titel :
Intelligent Information Systems,1994. Proceedings of the 1994 Second Australian and New Zealand Conference on
Conference_Location :
Brisbane, Qld.
Print_ISBN :
0-7803-2404-8
DOI :
10.1109/ANZIIS.1994.396938
