An approach to forecast red tide using generalized regression neural network

As a neural network provides a non-linear function mapping from input variables to the corresponding network output, without the requirements of having to specify the relation between the input and output variables in the form of mathematical formula, its widely used in modeling for complex non-linear phenomena. In this paper, generalized regression neural network (GRNN) is applied as a new type of model to forecast the red tide. Moreover, experiments with red tide data samples are performed in order to examine the usefulness of the method. Compared with the radial basis function (RBF) neural network, the experimental results are also analyzed.