An artificial intelligence approach for modeling the rejection of anti-inflammatory drugs by nanofiltration and reverse osmosis membranes using kernel support vector machine and neural networks

The rejection of anti-inflammatory drugs by membranes has shown paramount importance in separation membrane processes such as nanofiltration and reverse osmosis (NF/RO) membranes for pharmaceutical industries. Therefore, the main objective of this paper is to use support vector machine (SVM) and artificial neural network (ANN) to model the rejections of anti-inflammatory drugs by NF/RO membranes using 300 experimental data points gathered from the literature. Both approaches (ANN and SVM) gave close results with a slight superiority of the neural networks model demonstrated by its correlation coefficient (R) and root mean square error (RMSE) values of 0.9930 and 1.8094% respectively, in contrast to 0.9900 and 2.2355% for SVM. Sensitivity analysis by the weight method demonstrates that the most relevant variables that influence the rejection of anti-inflammatory drugs are: effective diameter of an organic compound in water "d(c) ", molecular length, contact angle, and zeta potential. These input relevant variables have a significant contribution (relative importance superior to 10%).