Wing Hong Chan, Albert W. M. Lee, Daniel W. J. Kwong, Yi-Zeng Liang and Ke-Min Wang
An optode device that contains two optical ion-selective membranes for potassium and sodium was used to obtain optode spectra of mixtures of these two ions. Such spectra are a non-linear function of the concentrations of the two ions. A back-propagation artificial neural network (BP-ANN) model was used to analyse these mixture optode spectra. The results showed that the BP-ANN technique is satisfactory for treating such non-linearity embedded in the data. In addition, an index for addressing the quality of the model approximation was defined and shown to be highly effective in controlling over-fitting in BP-ANN training.