LEARNING USING ERROR BACKPROPAGATION: A NEW VERSION | ||||
| International Conference on Aerospace Sciences and Aviation Technology | ||||
| Article 68, Volume 9, ASAT Conference, 8-10 May 2001, May 2001, Page 1-13 PDF (1.23 MB) | ||||
| Document Type: Original Article | ||||
| DOI: 10.21608/asat.2001.31148 | ||||
 View on SCiNiTO | ||||
| Authors | ||||
| MOUSSA H. M.1; ELARABY S. M.2; KOUTB M. A.3 | ||||
| 1Electronics Engineer, Atomic Energy Authority, Nuclear research Center, Anshas, Egypt. | ||||
| 2Associate Professor, Atomic Energy Authority, Nuclear research Center, Anshas, Egypt. | ||||
| 3Professor, Dept. of Industrial Electronics and Control, Faculty of Electronic Engineering, Menoufia, Egypt. | ||||
| Abstract | ||||
| ABSTRACT Training of multilayered feed-forward neural networks (MLFFNNs) is considered in this work. A procedure to derive high performance learning algorithm for updating the network weights is proposed. The proposed algorithm is based on heuristic technique that is developed from an analysis of the performance of the basic-backpropagation training algorithm. A unified formulation of the conventional learning algorithms including the basic-backpropagation algorithm, the momentum algorithm, and the exponential-smoothing algorithm alongside with the proposed learning algorithm is introduced. Recursive relations for updating the weights of the network are derived which greatly simplifies the application of these rules. Simulation results are presented and comparative studies are carried out to demonstrate the effectiveness of the new learning algorithm. The new algorithm can converge more than hundred times faster than the conventional algorithms. | ||||
| Keywords | ||||
| Neural Networks; Learning; Backpropagation and Convergence | ||||
|
Statistics Article View: 92 PDF Download: 178 |
||||
