Advantages of Artificial Neural Networks (ANNs) as alternative modelling technique for data sets showing non-linear relationships using data from a galenical study on a solid dosage form

Abstract

Artificial Neural Networks (ANN) methodology was used to assess experimental data from a tablet compression study showing highly non-linear relationships (i.e. measurements of ejection forces) and compared to classical modelling technique (i.e. Response Surface Methodology, RSM). These kinds of relationships are known to be difficult to model using classical methods. The aim of this investigation was to quantitatively describe the achieved degree of data fitting and predicting abilities of the developed models. The comparison between the ANN and RSM was carried out both graphically and numerically. For comparing the goodness of fit, all data were used, whereas for the goodness of prediction the data were split into a learning and a validation data set. Better results were achieved for the model using ANN methodology with regard to data fitting and predicting ability. All determined ejection properties were mainly influenced by the concentration of magnesium stearate and silica aerogel, whereas the other factors showed very much lower effects. Important relationships could be recognised from the ANN model only, whereas the RSM model ignored them. The ANN methodology represents a useful alternative to classical modelling techniques when applied to variable data sets presenting non-linear relationships.