Title :
A new backpropagation learning algorithm
Author :
Volovici, Daniel ; Mihalache, Adrian ; Iridon, Mihaela ; Breazu, Macarie
Author_Institution :
Sibiu Univ., Romania
Abstract :
Analyses and compares several improvements to the backpropagation method for weights adjustments for a feedforward network. The networks´ behavior is simulated on five specific applications. The paper also presents a method using a variable step and its superiority is proved by simulation
Keywords :
backpropagation; conjugate gradient methods; convergence; feedforward neural nets; backpropagation learning algorithm; network behavior; weights adjustments; Backpropagation algorithms; Biological neural networks; Biology computing; Computer errors; Computer networks; Error correction; Feedforward systems; Hardware; Mean square error methods; Software algorithms;
Conference_Titel :
Neural Networks, 1999. IJCNN '99. International Joint Conference on
Conference_Location :
Washington, DC
Print_ISBN :
0-7803-5529-6
DOI :
10.1109/IJCNN.1999.832640
