Title :
Comparison of gradient based training algorithms for multilayer perceptrons
Author :
Irwin, George ; Lightbody, Gordon ; McLoone, Sean
Author_Institution :
Dept. of Electr. & Electron. Eng., Queen´´s Univ., Belfast, UK
Abstract :
The training speed of batch backpropagation using steepest descent, conjugate gradient and quasi-Newton algorithm for a feedforward neural network are compared. Results illustrating the advantages of the Hessian based techniques are given and issues affecting speed discussed
Keywords :
Hessian matrices; Newton method; backpropagation; conjugate gradient methods; feedforward neural nets; multilayer perceptrons; Hessian based techniques; batch backpropagation; conjugate gradient; feedforward neural network; gradient based training algorithms; multilayer perceptrons; quasi-Newton algorithm; steepest descent;
Conference_Titel :
Advances in Neural Networks for Control and Systems, IEE Colloquium on
Conference_Location :
Berlin
