Assessment of competitive dye removal using a reliable method

Abdi, J ; Sharif University of Technology

763 Viewed
  1. Type of Document: Article
  2. DOI: 10.1016/j.jece.2014.06.002
  3. Abstract:
  4. In this study, a reliable and predictive model namely, least-squares support vector machine (LS-SVM) was developed to predict dye removal efficiency. Four LS-SVM models have been developed and tested using more than 630 series of experimental data which were obtained from our previous paper. These data consist of adsorbate type, adsorbent dosage, initial dye concentration, salt, absorbance time and dye removal efficiency. Direct Red 31 (DR31), Direct Green 6 (DG6) and Acid Blue (AB92) were used as a model dyes. The results show that the developed model is more accurate and reliable with the average absolute relative deviation of 0.678%, 0.877%, 0.581% and 0.978% for single systems and ternary system, respectively and correlation coefficients close to unity for all systems. Additionally, it is demonstrated that the proposed method is capable of simulating the actual physical trend of the dye removal efficiency with variation of adsorbent dosage and initial dye concentration in single and ternary systems. Eventually, the Leverage approach, in which the statistical Hat matrix, Williams plot, and the residuals of the model results lead to identification of the likely outliers, has been carried out. Fortunately, all the experimental data seem to be reliable except five in single systems. Therefore, the developed model could be reliable for prediction of the dye removal efficiency in its applicability domain. Selectivity analysis showed that GPN had selective removal of DG6
  5. Keywords:
  6. Selectivity analysis ; Ternary systems ; Average absolute relative deviations ; Competitive adsorption ; Correlation coefficient ; Dye removal ; Initial dye concentration ; Least squares support vector machines ; Leverage approach ; Predictive modeling ; Reliable methods ; Support vector machines
  7. Source: Journal of Environmental Chemical Engineering ; Vol. 2, issue. 3 , September , 2014 , p. 1672-1683
  8. URL: http://www.sciencedirect.com/science/article/pii/S2213343714001183