M Malekdar - Department of Chemical Engineering, Iran University of Science and Technology, Tehran, Iran
T Mohammadi - Department of Chemical Engineering, Iran University of Science and Technology, Tehran, Iran

Neural network models were used to predict crystallinity of synthesized ZSM-5 zeolite using different cations, Si/Al ratios and alkalinities in the starting gels at differentcrystallization times and temperatures. The ZSM-5 zeolite is used for producing a great number of membranes. Three neural network models with two hidden layers were constructed to predict the zeolite crystallinity.Effects of the input parameters on the crystallinity were also investigated using a limited number of training data