Please use this identifier to cite or link to this item:
|Title:||Hidden node activation differential - a new neural network relevancy criteria||Authors:||Hiang, Patrick Chan Khue
Erdogan, Sevki S.
|Issue Date:||1997||Citation:||Hiang, Patrick Chan Khue, Erdogan, Sevki S., Geok-See, Ng (1997). Hidden node activation differential - a new neural network relevancy criteria. International Conference on Knowledge-Based Intelligent Electronic Systems, Proceedings, KES 1 : 274-281. ScholarBank@NUS Repository.||Abstract:||Neural networks have been used in many problems such as character recognition, time series forecasting and image coding. The generalization of the network depends on its internal structure. Network parameters should be set correctly so that data outside the class will not be overfitted. One mechanism to achieve an optimal neural network structure is to identify the essential components (hidden nodes) and to prune off the irrelevant ones. Most of the proposed criteria used for pruning are expensive to compute and impractical to use for large networks and large training samples. In this paper, a new relevancy criteria is proposed and three existing criteria are investigated. The properties of the proposed criteria are covered in detail and their similarities to existing criteria are illustrated.||Source Title:||International Conference on Knowledge-Based Intelligent Electronic Systems, Proceedings, KES||URI:||http://scholarbank.nus.edu.sg/handle/10635/132910|
|Appears in Collections:||Staff Publications|
Show full item record
Files in This Item:
There are no files associated with this item.
checked on Nov 8, 2019
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.