Nonlinear Autoregressive Neural Network for Antimicrobial Waste Water Treatment

Abstract

Antibiotics become an emerging contaminant and receive more interests due to its ecotoxicological and strong stability in water ecosystems. Antibiotic adsorption onto carbon materials are biochars among the wastewater mechanisms. This research used machine learning (ML) techniques to generate general adsorption forecasting model for sulfamethoxazole (SMX) and tetracycline (TC) on CBM. Dirichlet design parameters and a combined combination of Neumann and Dirichlet boundary situation are applied to the system of differential equations. In addition, the proposed method use the learning under supervision technique of a nonlinear autoregressive for estimating the CO2 concentration and flows in units of rate of a reaction characteristics, an exogenous (NARX) neural network model with two activation functions was used (Log-sigmoid and hyperbolic tangent) and for both the findings of a TC and SMX absorption simulations showed the random forest performed support vector tree and nonlinear autoregressive exogenous neural networks and machine learning methods. Their relevance and complete dependency graph evaluation lead reasonable CBM uses for antimicrobial wastewater treatment. Also, machine learning forecasting model with good generalization capability is useful for building effective CBMs with few empirical screens. It evaluates the accuracy, precision, recall, false positive rate (FPR), and false negative rate (FNR) and also reduces the experimental screening.