Time Series Forecasting using CCA and Kohonen Maps - Applications to Electricity Consumption
(2000) European Symposium on Artificial Neural Networks (ESANN′2000) — Location: Bruges (Belgium) (26.April.2000)
Files
50-TimeseriesforecastingusingCCAandKohonenmaps-applicationtoelectricityconsumption.pdf
Restricted Access - Adobe PDF
- 1.7 MB
Details
- Authors
- Abstract
- Using large regressors in non-linear time series forecasting makes the fitting of the model difficult. This paper shows how to reduce the size of regressors in order to improve the forecasting performances, using the Curvilinear Component Analysis as projection tool. The method is applied to the Polish electrical load forecasting.
- Affiliations
Citations
Lendasse, A., Lee, J., Wertz, V., & Verleysen, M. (2000). Time Series Forecasting using CCA and Kohonen Maps - Applications to Electricity Consumption. Proceedings of ESANN 2000, European Symposium on Artificial Neural Networks, p. 329-334. https://hdl.handle.net/2078.5/254052