Artificial Neural Network (ANN) approach for modeling experimental data of viscosity and density of a ternary solution

The Artificial neural network (ANN) approach was used to model experimental viscosity anddensity data for ternary aqueous solutions of calcium chloride and potassium chloride. Then, the ANNmodel was compared with the previously investigated models e.g., modified Jones-Dole, Hu,Exponential and GF models used for the same dataset. In the present study, the Levenberg-Marquardtalgorithm or "trainlm" command was selected as the training algorithm. Subsequently, differentconfigurations of the network were compared and the optimal multi-layer perceptron (MLP) networkwas designed with ۳ hidden layers and [۸ ۴ ۳] neurons since it showed better performance. Moreover,۸۰% of the dataset for network training, ۱۰% for validation and the rest for network testing wererandomly selected. Amid investigated models, ANN obtained the minimum mean square error (MSE) of۶.۲۰۰۸×۱۰-۵ and maximum R۲ of ۰.۹۹۹۷ while at best modifies Jones-Dole could achieve a MSE of۲.۷۶۸×۱۰-۵ and R۲ of ۰.۹۹۹۶. suggesting that the ANN model is the most optimal model for modelingthe viscosity and density of this ternary solution.

Keywords:

Artificial neural network (ANN) , Viscosity and density , Ternary solutions , multi-layerperceptron (MLP) ,