摘要：An Artificial Neural Network (ANN) model was urbanized to forecast the biosorption competence of zincoxide nanoparticle ingrained on activated silica using Corriandrum sativum (ZNO-NPs-AS-Cs) for theamputation of whole As(III) from aqueous solution based on 95 data sets obtained in a laboratory batchstudy. Experimental parameters affecting the biosorption progression such as initial concentration, dosage,pH, contact time and agitation were premeditated. A contact time of 90 min was generally passable to bringabout equilibrium. The maximum adsorption capacity of (ZNO-NPs-AS-Cs) in AS (III) removal was found tobe 3.46 g/L. The sensitivity analysis confirmed that MSE values decreased as the number of variables usedin the ANN model increased. The relative increase in the performance due to inclusion of V2, adsorbentdosage; V3, contact time; and V5, agitation speed is larger than the contribution of other variables. Theproposed ANN model provided realistic experimental data with a satisfactory correlation coefficient of 0.999for five operating variables..
