ENHANCEMENT SPECTRAL RESOLUTION FOR THE PREDICTION AMOUNT OF SOFOSBUVIR AND LEDIPASVIR USING LEAST SQUARES SUPPORT VECTOR MACHINE AND ARTIFICIAL NEURAL NETWORKS IN PHARMACEUTICAL FORMULATION

Abstract

In this study, least squares support vector machine (LS-SVM) and artificial neural networks (ANNs) as intelligent methods combined with spectrophotometry method, were used for determination of Sofosbuvir (SOF) and Ledipasvir (LED) in synthetic mixtures and Harvoni tablet simultaneously. In the LS-SVM method, Radial Basis Function (RBF) was used as kernel function. Then, the regularization parameter (γ) and Bandwidth (2) were optimized and root mean square error prediction (RMSE) was 0.4164, 0.6033 for SOF and LED respectively. Afterwards, Feed-forward back-propagation network with different training algorithms was used in artificial neural network method. These training algorithms compared with each other for selecting the best model. On the other hand, radial basis function neural network (RBFNN) was applied as an efficient network. Finally, these methods were compared to the high- performance liquid chromatography (HPLC) as a reference method. According to one way analysis of variance (ANOVA) test at the 95 % confidence level, there were no significant differences between LS-SVM, ANN and reference methods.