Evolving time-lagged feedforward neural networks for time series forecasting

• Autor(es): Peralta Donate, Juan ; Cortez, Paulo, 1971- ; Gutierrez Sanchez, German ; Sanchis de Miguel, Araceli
• Data: 2011
• Identificador Persistente: http://hdl.handle.net/1822/14849
• Origem: RepositóriUM - Universidade do Minho
• Assunto(s): Connectionism and neural nets; Hybrid systems
• Tipo de Documento: Documento de conferência
• Idioma: Inglês

Descrição

Time Series Forecasting (TSF) is an important tool to sup- port both individual and organizational decisions. In this work, we propose a novel automatic Evolutionary Time- Lagged Feedforward Network (ETLFN) approach for TSF, based on an Estimation Distribution Algorithm (EDA) that evolves not only Artificial Neural Network (ANN) parame- ters but also which set of time lags are fed into the fore- casting model. Such approach is compared with similar strategy that only selects ANN parameter and the conven- tional TSF ARIMA methodology. Several experiments were held by considering six time series from distinct domains. The obtained multi-step ahead forecasts were evaluated us- ing SMAPE error criteria. Overall, the proposed ETLFN method obtained the best forecasting results. Moreover, it favors simpler neural network models, thus requiring less computational effort.