Simultaneous Prediction of the Density, Viscosity and Electrical Conductivity of Pyridinium-Based Hydrophobic Ionic Liquids Using Artificial Neural Network

Abstract

An Artificial Neural Network (ANN) model is presented for the Simultaneous prediction of density, viscosity and electrical conductivity of pyridinium-based hydrophobic ionic liquids. Data density, viscosity and electrical conductivity obtained from paper and from a three layer feed forward artificial neural network is used to estimate them. The optimal ANN model consisted of, two neurons in the input layer, ten neurons in the hidden layer and three neurons in the output layer. This model predicts the density with a Mean Square Error (MSE) of 7.5714 × 10− 7 and the coefficient of determination (R2) of 1.0000, viscosity with a Mean Square Error (MSE) of 1.1332 × 10− 4 and the coefficient of determination (R2) of 0.9982 and electrical conductivity with a Mean Square Error (MSE) of 2.2668 × 10− 6 and the coefficient of determination (R2) of 0.9999. The results show that the Simultaneous predicted of density, viscosity and electrical conductivity of pyridinium-based hydrophobic ionic liquids by using artificial neural network well done. The artificial neural network model shows lower errors and higher precision compared to statistical models while use of ANN is easier and quicker than statistical methods.