Wavelet-artificial neural network to predict the acetone sensing by indium oxide/iron oxide nanocomposites

Comparison of hybrid RNA-based models to predict the degradation and mineralization of the microcontaminant hormone 17β-estradiol

New alternatives for effluent decontamination, such as electrochemical oxidation, are being developed to provide adequate removal of endocrine disruptors such as 17β-estradiol in wastewater. In this study, data-driven models of response surface methodology, artificial neural networks, wavelet neural networks, and adaptive neuro-fuzzy inference system will be used to predict the degradation and mineralization of the microcontaminant hormone 17β-estradiol through an electrochemical process to contribute to the treatment of effluent containing urine. With the use of different statistical criteria and graphical analysis of the correlation between observed and predicted data, it was possible to conduct a comparative analysis of the performances of the data-driven approaches. The results point to the superiority of the adaptive neuro-fuzzy inference system (correlation coefficient, R2, ranged from 0.99330 to 0.99682 for TOC removal and from 0.95330 to 0.99223 for the degradation of the hormone 17β-estradiol) techniques over the others. The remaining results obtained with the other metrics are consistent with this analysis.

Applying feature selection and machine learning techniques to estimate the biomass higher heating value

The biomass higher heating value (HHV) is an important thermal property that determines the amount of recoverable energy from agriculture byproducts. Precise laboratory measurement or accurate prediction of the HHV is essential for designing biomass conversion equipment. The current study combines feature selection scenarios and machine learning tools to establish a general model for estimating biomass HHV. Multiple linear regression and Pearson's correlation coefficients justified that volatile matter, nitrogen, and oxygen content of biomass samples have a slight effect on the HHV and it is better to ignore them during the HHV modeling. Then, the prediction performance of random forest, multilayer and cascade feedforward neural networks, group method of data handling, and least-squares support vector regressor are compared to determine the intelligent estimator with the highest accuracy toward biomass HHV prediction. The ranking test shows that the multilayer perceptron neural network better predicts the HHV of 532 biomass samples than the other intelligent models. This model presents the outstanding absolute average relative error of 2.75% and 3.12% and regression coefficients of 0.9500 and 0.9418 in the learning and testing stages. The model performance is also superior to a recurrent neural network which was recently developed in the literature using the same databank.