Development of artificial neural network on transparent soap base containing Sonneratia caseolaris extract

Abstract

The objective of this study was to use artificial neural network in development of transparent soap. The different eighteen transparent soap formulations were prepared and the physical properties of them such as clearness, hardness, foam ability and surface tension were investigated. Moreover, the correlation between each formulation and response parameters was examined using feed-forward back-propagation neural networks. The results showed that the amounts of SLES-N70, glycerine, sodium stearate and PVP-K30 were the important parameters on foam ability, clearness, hardness and surface tension, respectively. The proposed models were able to predict the properties of transparent soap with a reasonable degree of accuracy. The predictive ability of these models was validated by an external set of 6 formulations which were not included in the training set. The predictions were in good agreement with the observed and the predictived values. Moreover, the 5% of Sonneratia caseolaris extract was successfully incorporated into the soap. These results could be applicable for development of transparent soap containing S. caseolaris extract.