ABSTRACT
Artificial neural networks (NN) can be applied in the form of stand-alone models, or combined with other tools including other NN, to provide hybrid modelling solutions. Hybrid neural network (HNN) models can be defined as the integration of a number of different models, but with the proviso that one or more of the constituent models is a NN. The underlying principle of the hybrid model is that it exploits the strength of the individual component models in a synergistic manner to produce a better forecasting solution. The hybrid model offers opportunities for integrating conventional hydrological models with those based on artificial intelligence techniques, such as neurocomputing and fuzzy logic. Moreover, in hybrid modelling, conventional models and the artificial intelligence solutions are intended to complement rather than compete with each other.
