• We are available for your help 24/7
  • Email: info@isindexing.com, submission@isindexing.com

Paper Details

Alternative method to predict activity coefficients at infinite dilution of hydrocarbons in aqueous solutions

Gonzalo Astray, Jorge Morales, Miriam Gonzlez-Temes, Juan C. Mejuto, Augusto J. Magdalena

Journal Title:Mediterranean Journal of Chemistry

Activity coefficients at infinite dilution are important property in solute-solvent interactions. Experimental techniques show high costs, skilled labour and safety. To solve this, a neural network model with five different topological descriptors to implement different Artificial Neural Network has been implemented. The best Artificial Neural Network (5-11-8-1 topology) presents good fits for the training phase with an Average Percentage Deviation of 1.85%. Similar results have been obtained for the validation phase of the Artificial Neural Network (1.88%). The implemented Neural Networks techniques showed better results than other developed methods, around 30.70% and 24.60% for training and validation phase, respectively.