Forecasting using genetic programming
Alaa F. Sheta, Ahmed Khairy Mahmoud · 2002
In this paper, two models for forecasting the Nile River flow have been developed. The traditional linear autoregressive (AR) model and genetic programming (GP) based model are presented. The performance of both the AR and GP models were tested using a set of measurements recorded at the Donagola station located in the Northern Sudan. A significant improvement of the error when using the GP model for forecasting was achieved.