Comparison different neural machine learning models for predicting of nano-size pharmaceuticals

In this paper, we investigate the applicability of MATLAB to design, train and test an artificial neural network (ANN) to determine the relationships of variables in drug nanoprecipitation using microfluidic reactor. Effective variables on nanoparticle size are computed as ANN inputs and the particle size is considered to be the output. After the training on the input–output process, the AAN predicted values were compared with the other available AAN model which is obtained from INForm software (INForm v3.5, Intelligensys, UK) [1]. Comparing coefficient of determination (R2), regression coefficient (R) values of the MATLAB-ANN model and INForm-ANN model, one can conclude that the first model predicted the particle size with more accuracy than the other.