Paulo S. G. de Mattos Neto,  Gustavo G. Petry,  João P. de M. Ataide,  Tiago A. E. Ferreira. 

In this work is presented a hybrid approach by the composition of a modified genetic algorithm and MLP artificial neural networks. The methodology is a simple form to carry out experiments based on artificial neural networks when compared to the conventional form of the network experiments (choice and adjustment of the parameters). Especially, the developed methodology is applied to the time series forecast problem. Experimental results are showed with a set of four relevantes time series.

http://www.lbd.dcc.ufmg.br/colecoes/enia/2005/068.pdf

Processando consulta.