Title :
Variance analysis of sensitivity information for pruning multilayer feedforward neural networks
Author :
Engelbrecht, AP ; Fletcher, L. ; Cloete, I.
Author_Institution :
Pretoria Univ., South Africa
Abstract :
This paper presents an algorithm for pruning feedforward neural network architectures using sensitivity analysis. Sensitivity Analysis is used to quantify the relevance of input and hidden units. A new statistical pruning heuristic is proposed, based on the variance analysis, to decide which units to prune. Results are presented to show that the pruning algorithm correctly prunes irrelevant input and hidden units
Keywords :
feedforward neural nets; optimisation; sensitivity analysis; statistical analysis; feedforward neural networks; heuristic; multilayer neural networks; pruning; sensitivity analysis; statistical analysis; variance analysis; Africa; Analysis of variance; Feedforward neural networks; Information analysis; Information technology; Multi-layer neural network; Neural networks; Robust stability; Sensitivity analysis; Statistical analysis;
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.1999.832657
