Skip to main content
placeholder image

Novel training algorithm based on quadratic optimisation using neural networks

Chapter


Abstract


  • In this paper we present a novel algorithm for training feedforward neural networks based on the use of recurrent neural networks for bound constrained quadratic optimisation. Instead of trying to invert the Hessian matrix or its approximation, as done in other second-order algorithms, a recurrent equation that emulates a recurrent neural network determines the optimal weight update. The development of this algorithm is presented, along with its performance under ideal conditions as well as results from training multilayer perceptrons. The results show that the algorithm is capable of achieving results with less errors than other methods for a variety of problems. © Springer-Verlag Berlin Heidelberg 2001.

Publication Date


  • 2001

Citation


  • Arulampalam, G., & Bouzerdoum, A. (2001). Novel training algorithm based on quadratic optimisation using neural networks. In Unknown Book (Vol. 2084 LNCS, pp. 410-417). doi:10.1007/3-540-45720-8_48

International Standard Book Number (isbn) 13


  • 9783540422358

Scopus Eid


  • 2-s2.0-84902134920

Web Of Science Accession Number


Book Title


  • Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Start Page


  • 410

End Page


  • 417

Abstract


  • In this paper we present a novel algorithm for training feedforward neural networks based on the use of recurrent neural networks for bound constrained quadratic optimisation. Instead of trying to invert the Hessian matrix or its approximation, as done in other second-order algorithms, a recurrent equation that emulates a recurrent neural network determines the optimal weight update. The development of this algorithm is presented, along with its performance under ideal conditions as well as results from training multilayer perceptrons. The results show that the algorithm is capable of achieving results with less errors than other methods for a variety of problems. © Springer-Verlag Berlin Heidelberg 2001.

Publication Date


  • 2001

Citation


  • Arulampalam, G., & Bouzerdoum, A. (2001). Novel training algorithm based on quadratic optimisation using neural networks. In Unknown Book (Vol. 2084 LNCS, pp. 410-417). doi:10.1007/3-540-45720-8_48

International Standard Book Number (isbn) 13


  • 9783540422358

Scopus Eid


  • 2-s2.0-84902134920

Web Of Science Accession Number


Book Title


  • Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Start Page


  • 410

End Page


  • 417