Adaptive multiresolution filtering to forecast nonlinear time series

E. Gómez-Ramirez, X. Vilasis-Cardona

Producció científica: Contribució a una conferènciaContribucióAvaluat per experts

Resum

There are two ways to improve the Identification Process of a dynamic system using an Artificial Neural Network: 1. Preprocessing the training values to extract characteristics of the data, and 2. Adapting the architecture of the network. In this paper we used an Adaptive Scheme of Multiresolution Filtering to decompose the series into other series easier to analyze. The scheme proposed uses Genetic Algorithm (GA) to find the optimal bank of filters without previous knowledge of the behavior of the system to be identified. A new variation of the algorithm using random individuals is proposed to avoid local minima. The objective function proposed is the estimation quadratic error of a Multilayer Perceptron using Levenberg-Maquardt Learning.

Idioma originalAnglès
Pàgines400-405
Nombre de pàgines6
Estat de la publicacióPublicada - 2002
Esdeveniment2002 International Joint Conference on Neural Networks (IJCNN '02) - Honolulu, HI, United States
Durada: 12 de maig 200217 de maig 2002

Conferència

Conferència2002 International Joint Conference on Neural Networks (IJCNN '02)
País/TerritoriUnited States
CiutatHonolulu, HI
Període12/05/0217/05/02

Fingerprint

Navegar pels temes de recerca de 'Adaptive multiresolution filtering to forecast nonlinear time series'. Junts formen un fingerprint únic.

Com citar-ho