Phytoextraction capability of Azolla pinnata in the removal of rhodamine B from aqueous solution: artificial neural network and random forests approaches

Abstract

This study used live Azolla pinnata (AP) to remediate rhodamine B (RB) from aqueous solutions via the phytoextraction method, and machine learning algorithms such as artificial neural networks and random forests were used as predictive models. The pH was found to have a major influence on the phytoextraction process, and the AP dosage can change the pH of the aqueous solution. The optimum condition for the phytoextraction of RB (initial dye concentration at 10 ppm) is at pH 3.0 with a plant dosage of 0.4 g, resulting in removal efficiency as high as 76%. The growth estimation (relative frond number) indicates that AP can tolerate RB concentration of as high as 20 mg L −1 , and the estimations from the pigment studies showed that the exposure of AP to RB causes AP to produce higher concentrations of plant pigments than the control, which hinted the possibility of AP using RB and its intermediates for growth.

A unified framework of response surface methodology and coalescing of Firefly with random forest algorithm for enhancing nano-phytoremediation efficiency of chromium via in vitro regenerated aquatic macrophyte coontail (Ceratophyllum demersum L.)

“…Br.) also presented the R 2 and MSE scores for the removal of malachite green (Kooh et al 2016 ). The use of ML modeling for phytoremediation of heavy metals in soils has already been documented (Shi et al 2023 ) like immobilization efficiency in biochar-amended soils (Palansooriya et al 2022 ), and Cd removal by Sinapas alba L. (Jaskulak et al 2020 ).…”

Dye Pollutant Removal from Synthetic Wastewater: A New Modeling and Predicting Approach Based on Experimental Data Analysis, Kriging Interpolation Method, and Computational Intelligence Techniques

“…Thus, modeling and optimization of wastewater treatment processes utilizing artificial intelligence (AI) due to their unique advantages (like a low number of parameters, low computation time and no need for boundary and initial condition), with a focus on achieving the maximum removal efficiency of various pollutants, especially dye pollutants, has received much attention (Fan et al, 2018;Mossavi et al, 2019). Some papers investigated the performance of this method in modeling of different wastewater treatment processes such as using artificial neural network (ANN) with 54 samples for adsorption of Methyl Orange (MO) (Tanhaei et al, 2016), Acid Red 33 electrochemical decolorization modeling by ANN with 78 samples (Chianeh et al, 2017), modeling of Rhodamine B remediation from aqueous solutions via the phytoextraction method by ANN (with 154 samples) and random forest (RF) (Kooh et al, 2019), utilizing of support vector machine (SVM) with 249 samples for modeling the photocatalytic degradation of methyl tert-butyl ether (Oyehan et al, 2019), reduction of phosphorus modeling by linear regression, ANN, and M5P with 106 number of data (Kumar and Deswal, 2020), modeling of Pb +2 adsorption by ANN and multivariate linear regression (MLR) using 20 samples (Ashrafi et al, 2020), and adsorption Pb(II), Ni(II), and Cu(II) modeling through ANN with 476 samples (Hanandeh et al, 2021).…”

Application of multilayer perceptron network and random forest models for modelling the adsorption of chlorobenzene on a modified bentonite by intercalation with hexadecyltrimethyl ammonium (HDTMA)

“…Therefore, in this study, our prime focus is on two algorithms, namely, artificial neural network (ANN) and random forest (RF). Recently, these algorithms have become quite popular in machine learning community due to their robustness, ability to generalize well on unseen data, ease of use and, implementation on real-world problems [35,36]. The main aim of this work is to investigate the removal of CB from aqueous solution using HDTMA modified bentonite as an adsorbent.…”