Abstract
A novel hybrid algorithm based on the AFTER (Aggregated forecast through exponential re-weighting) and the modified particle swarm optimization (PSO) is proposed. The combining weights in the hybrid algorithm are trained by the modified PSO. The linear constraints are added in the PSO to ensure that the sum of the combining weights is equal to one. Simulated results on the prediction of the stocks data show the effectiveness of the hybrid algorithm. © Springer-Verlag Berlin Heidelberg 2004.
Cite
CITATION STYLE
Feng, X., Liang, Y., Sun, Y., Lee, H. P., Zhou, C., & Wang, Y. (2004). A hybrid algorithm for combining forecasting based on AFTER-PSO. In Lecture Notes in Artificial Intelligence (Subseries of Lecture Notes in Computer Science) (Vol. 3157, pp. 942–943). Springer Verlag. https://doi.org/10.1007/978-3-540-28633-2_105
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
