Linear Polyethers as Additives for AOT-Based Microemulsions: Prediction of Percolation Temperature Changes Using Artificial Neural Networks

Abstract

Predictive models based on artificial neural networks have been developed for the percolation threshold of AOT based microemulsions with addition of either glymes or polyethylene glycols. Models have been built according to the multilayer perceptron architecture, with five input variables (concentration, molecular mass, log P, number of C and O of the additive). Best model for glymes has a topology of five input neurons, five neurons in a single hidden layer and one output neuron. Polyethylene glycol model's architecture consists in five input neurons, three hidden layers with eight neurons in both first two and five in the last, and a neuron in the last output layer. All of them have a good predictive power according to several quality parameters.

Influence of Amphiphiles on Percolation of AOT-Based Microemulsions Prediction Using Artificial Neural Networks

“…The best developed neural model to predict glymes percolation temperature presents a topology 5-5-1, that is, five nodes in input layer, five nodes in the only intermediate layer and one neuron in the output layer [55]. This neural model has been trained with 32 experimental cases, and 11 experimental cases were used to validate the neural model [55].…”

“…Figure 7 shows a scheme of this neural model [55]. Best polyethylene glycols model presents a topology with five input neurons, three intermediate layers with eight, eight and five neurons and an output layer with one node (see Figure 8) [55]. This model was developed using 68 training cases and were validated with 14 experimental cases [55].…”

“…This neural model has been trained with 32 experimental cases, and 11 experimental cases were used to validate the neural model [55]. Figure 7 shows a scheme of this neural model [55].…”

“…This neural model has been trained with 32 experimental cases, and 11 experimental cases were used to validate the neural model [55]. Figure 7 shows a scheme of this neural model [55]. Best polyethylene glycols model presents a topology with five input neurons, three intermediate layers with eight, eight and five neurons and an output layer with one node (see Figure 8) [55].…”

“…Available datasets for glymes [55] consisted of 44 microemulsion compositions and for polyethylene glycols [55][56][57] consisted of 82 microemulsion compositions.…”