Comparison of regression and design models for biosorption process

Abstract

In this study, the biosorption process were optimized and compared for the first time in terms of their accuracy and predictive ability for the sorption of Cu(II) ions onto date palm (Phoenix dactylifera L.) seeds used as an agricultural waste product by using four models, multi-linear regression (MLR), full factorial design with center points (FFD), Box-Behnken design (BBD) and central composite design (CCD). The responses were evaluated based on the regression equations formulated according to the results of the analyses of models. It was found that MLR and FFD models had a lower predictive capability than response surface methodologies (RSM).