Artificial neural networks prediction for blood concentration and dosage of cyclosporine A

This paper used methods of multiple linear regression (MLR), back propagation artificial neural network (BPANN) and genetic algorithm optimized back propagation artificial neural network (GA-BPANN) to predict blood concentration and dosage of cyclosporine A.It is proved that GABPNN model predict CsA blood concentration or dosage more accurate than MLR model or BPANN model by using 10-fold cross-validation. Besides, GABPANN model is more stable and reasonable. The scheme of two chained GABPANN models can be efficiently applied to prediction of CsA blood concentration and dosage.