Mathematical modeling of liquid phase equilibria for the systems (water + phosphoric acid + linear alcohols) using the GMDH-type neural network

A GMDH type-neural network was applied to predict ternary equilibrium data for the {water + phosphoric acid + linear alcohol (1-nonanol or 1-undecanol)} ternary systems at T = 298.2 K. In order to achieve modeling, the experimental points were divided into train and test parts. The data set was separated into two sections: 70% were utilized for ‘‘training’’ and 30% were applied as test. The forecasted values were compared to experimental values in order to evaluate the ability of the GMDH neural network method. The results predicted by GMDH type neural network were in excellent agreement with the experimental results