COMPARISON OF PREDICTION POWER BETWEEN THEORETICAL AND NEURAL-NETWORKMODELS IN ION-INTERACTION CHROMATOGRAPHY

The separation by ion-interaction chromatography (IIC) of metal comple xes having single and double charges has been studied in order to comp are the prediction power of oft (neural-network) and hard modelling (I IC equation). The two approaches have been used to model the retention behaviour as a function of the composition of the mobile phase, With ion-interaction mobile phases, the parameters involved included the co ncentrations of ion-interaction reagent, organic modifier and ionic st rength. From a set of 69 experimental design points (the different mob ile phase compositions at which capacity factors are measured), one te st set of ten design points and ten training sets, containing from 59 to 11 design points, have been extracted. Chromatographic and chemomet ric considerations for the selection of the data sets and minimum numb er of observations required have been discussed. The study showed that the IIC equation predicted more accurately when few experimental data were available, while a similar prediction power was obtained with bo th models when the number of data was more than 17. Nevertheless the n eural-network accounted for a greater versatility without the need to develop an equation. (C) 1998 Elsevier Science B.V.