Prediction of two-metal biosorption equilibria using a neural network

A feedforward neural network model with a single hidden layer was used to correlate and predict biosorption equilibrium data in a binary metal system. Experimental data on the biosorption of Fe(III) and Cr(VI) by the microalga Chlorella vulgaris reported in the literature was used to assess the performance of the neural network. It was demonstrated that the neural network approach was significantly more accurate than the traditional modeling approach based on Langmuir-type models. To assess the predictive capability of the neural network model, the network was trained using a subset of available data. The suitably trained neural network was found to be capable of predicting fresh data not belonging to the training set. However, training data should be selected carefully if the best results are to be achieved. © 2003 SDU. All rights reserved.