Reply to “Quantitative structure–affinity relationship study of azo dyes for cellulose fibers by multiple linear regression and artificial neural network”

Abstract

A quantitative structure–property relationship study was recently published by Wang et al. (Chemom. Intell. Lab. Syst. 134 (2014) 1–9), to correlate 51 azo dye affinities for cellulose to dye molecular descriptors by multiple linear regression (MLR) and artificial neural network (ANN). They performed several tests to check the robustness and predictive power of their MLR and ANN models and compared their results with others previously published. Some erroneous calculations from their paper questioned the value of these comparisons. In this paper corrections on their published results are presented, and also additional metrics for model validation are added.