Simple spectrophotometric method for simultaneous determination of salmeterol and fluticasone as anti-asthma drugs in inhalation spray based on artificial neural network and support vector regression

Abstract

In this study, the application of artificial neural network (ANN) and support vector regression (SVR) methods with spectrophotometry approach was assessed for the simultaneous determination of salmeterol (SMT) and fluticasone (FLU) in synthetic mixtures and inhalation anti-asthma spray. Levenberg–Marquardt (LM) and gradient descent with adaptive learning rate backpropagation (GDA) were applied as training algorithms of feedforward neural network (FFNN). According to the results of mean square error (MSE), the LM algorithm had better ability than GDA for the prediction. Root mean square error (RMSE) of the LM algorithm related to the training, validation, and test sets were 0.168, 0.238, 0.192 and 0.275, 0.360, 0.331 for SMT and FLU, respectively. Also, the mean recovery of these sets was obtained between 99.11% and 102.39% for both components. Optimum parameter values of SVR model were found with minimum RMSE of 0.1068 and 0.1443 for SMT and FLU, respectively. In addition, the mean recovery of test set was achieved 100.91% and 100.46% for SMT and FLU, respectively. The analysis results of anti-asthma spray were compared with high-performance liquid chromatography (HPLC) as a reference technique. No significant difference was observed between these methods using one-way analysis of variance (ANOVA).