DocumentCode :
1842982
Title :
Sparse algorithm for feed-forward neural networks
Author :
Li, Wangchao ; Yongbin Yang ; Zhang, Jie ; Wang, Liying
Volume :
3
fYear :
1999
fDate :
1999
Firstpage :
1805
Abstract :
In this paper self-adjustment pruning algorithm is presented to make arbitrary feed-forward neural networks sparse. To adapt to the complex and flexible architecture of arbitrary feed-forward neural network, partial order and topology order are introduced. Based on the exemplar vector, discrete degree and similarity degree are presented. In order to search redundant neurons and connections, the criterions of redundant neurons and connections are brought up. In order to keep up the original input-output behavior it is required to modify the remaining weights and threshold locally, and create new connections sometimes. The change to the related weights and thresholds can be obtained by constructing and solving a system of linear equations. Experimental results show this algorithm is very effective and feasible. The proposed algorithm is not only faster than general sparse algorithm but also can keep up the original performance and improve the generalization. So it is not required to retrain the neural network after pruning
Keywords :
computational complexity; feedforward neural nets; generalisation (artificial intelligence); redundancy; self-organising feature maps; I/O behavior; complex flexible architecture; discrete degree; exemplar vector; feed-forward neural networks; input-output behavior; linear equations; partial order; redundant connection search; redundant neuron search; self-adjustment pruning algorithm; similarity degree; sparse algorithm; sparse feedforward neural networks; topology order; Equations; Feedforward neural networks; Feedforward systems; Feeds; Network topology; Neural networks; Neurons; Radio frequency;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Neural Networks, 1999. IJCNN '99. International Joint Conference on
Conference_Location :
Washington, DC
ISSN :
1098-7576
Print_ISBN :
0-7803-5529-6
Type :
conf
DOI :
10.1109/IJCNN.1999.832652
Filename :
832652
Link To Document :
بازگشت