Avaliação estatística de métodos de poda aplicados em neurônios intermediários da rede neural MLP
Contribuinte(s) |
Universidade Estadual Paulista (UNESP) |
---|---|
Data(s) |
27/05/2014
27/05/2014
01/12/2006
|
Resumo |
There are several papers on pruning methods in the artificial neural networks area. However, with rare exceptions, none of them presents an appropriate statistical evaluation of such methods. In this article, we proved statistically the ability of some methods to reduce the number of neurons of the hidden layer of a multilayer perceptron neural network (MLP), and to maintain the same landing of classification error of the initial net. They are evaluated seven pruning methods. The experimental investigation was accomplished on five groups of generated data and in two groups of real data. Three variables were accompanied in the study: apparent classification error rate in the test group (REA); number of hidden neurons, obtained after the application of the pruning method; and number of training/retraining epochs, to evaluate the computational effort. The non-parametric Friedman's test was used to do the statistical analysis. |
Formato |
249-256 |
Identificador |
http://dx.doi.org/10.1109/TLA.2006.4472121 IEEE Latin America Transactions, v. 4, n. 4, p. 249-256, 2006. 1548-0992 http://hdl.handle.net/11449/69266 10.1109/TLA.2006.4472121 2-s2.0-77958181102 2-s2.0-77958181102.pdf |
Idioma(s) |
por |
Relação |
IEEE Latin America Transactions |
Direitos |
openAccess |
Palavras-Chave | #Artificial Neural Network #Classification error rate #Classification errors #Computational effort #Experimental investigations #Hidden layers #Hidden neurons #Intermedia #MLP neural networks #Multilayer perceptron neural networks #Non-parametric #Pruning methods #Statistical analysis #Statistical evaluation #Function evaluation #Neural networks |
Tipo |
info:eu-repo/semantics/article |