Title :
Dynamic pruning algorithms for improving generalisation of neural networks
Author :
Babri, H.A. ; Kot, A.C. ; Tan, N.T. ; Tang, J.G.
Author_Institution :
Sch. of Electr. & Electron. Eng., Nanyang Technol. Inst., Singapore
Abstract :
In this paper, we study the effects of pruning on the generalisation performance of feed-forward networks. The weighted sensitivity method and ratio dependent sensitivity method are proposed to efficiently prune a given network. Pruning criteria using statistical methods are also discussed and experimental results are presented to illustrate their effects on the generalisation property of the network
Keywords :
feedforward neural nets; generalisation (artificial intelligence); feed-forward networks; generalisation; pruning; ratio dependent sensitivity; weighted sensitivity; Cost function; Feedforward systems; Heuristic algorithms; Integral equations; Interpolation; Neural networks; Neurons; Redundancy; Signal processing algorithms; Statistical analysis;
Conference_Titel :
Information, Communications and Signal Processing, 1997. ICICS., Proceedings of 1997 International Conference on
Print_ISBN :
0-7803-3676-3
DOI :
10.1109/ICICS.1997.652063
