The application of the Radial Basis Function Neural Network in estimation of nitrate contamination in Manawatu river

The Radial Basis Function (RBF) Neural Network has shown its strong capability in pattern recognition, classification and function approximation problems. In this paper, the RBF neural network is used to classify different levels of nitrate contamination in river water. The planar electromagnetic sensors have been subjected to different water samples contaminated by nitrate and output signals have been extracted. These signals are derived and its suitable features are extracted by using three different features; energy, mean and skewness. These features are inputted to the RBF neural network consequently, for the classification of different levels of nitrate concentration in water. The result shows that the planar electromagnetic sensor with the assistance of the RBF neural network can be a good alternative to current laboratory testing methods.