There are few methods to predict ion removal using ion flotation without lengthy experiments. The
objective of this study was to model the Zn(II) flotation using a hybrid neural network/genetic
algorithm (GANN) and multivariate linear regression (MLR). Mean square error and correlation
coefficient values of 0.9228 and 190.1, respectively, for testing the datasets of the GANN model reveal
the superiority of the GANN model in predicting the Zn(II) removal, while these values were obtained
as 0.9125 and 220.36, respectively, for the MLR model. The results showed that this model could be
useful in predicting the Zn(II) removal in response to changes in the input parameters.
