A Model Based on Bootstrapped Neural Networks for Modeling the Removal of Organic Compounds by Nanofiltration and Reverse Osmosis Membranes

Ammi, Yamina; Khaouane, Latifa; Hanini, Salah

doi:10.1007/s13369-018-3484-8

Cited by 9 publications

(9 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this work, we have used data available in the literature. [4][5][6][7][8]14,[17][18][19][29][30][31][32][33][34][35][36][37][38][39][40][41][42][43][44][45][46] The first database (DB 1 ) consists of 1394 rejections of 116 organic compounds, the second database (DB 2 ) consists of 980 rejections of 102 organic compounds, and the third database (DB 3 ) consists of 436 rejections of 42 organic compounds. The list of the 116 organic compounds is shown in Supplementary Data A.…”

Section: Data Base Collectionmentioning

confidence: 99%

A Comparison of “Neural Networks and Multiple Linear Regressions” Models to Describe the Rejection of Micropollutants by Membranes

2020

Self Cite

View full text Add to dashboard Cite

A rejection process of organic compounds by nanofiltration and reverse osmosis membranes was modelled using the artificial neural networks. Three feed-forward neural networks based on quantitative structure-activity relationship (QSAR-NN models) characterised by a similar structure (twelve neurons for QSAR-NN1, QSAR-NN2, and QSAR-NN3 in the input layer, one hidden layer and one neuron in the output layer), were constructed with the aim of predicting the rejection of organic compounds. A set of 1394 data points for QSAR-NN1, 980 data points for QSAR-NN2, and 436 data points for QSAR-NN3 were used to construct the neural networks. Good agreements between the predicted and experimental rejections were obtained by QSAR-NN models (the correlation coefficient for the total dataset were 0.9191 for QSAR-NN1, 0.9338 for QSAR-NN2, and 0.9709 for QSAR-NN3). Comparison between the feed-forward neural networks and multiple linear regressions based on quantitative structure-activity relationship “QSAR-MLR” revealed the superiority of the QSAR-NN models (the root mean squared errors for the total dataset for the QSAR-NN models were 10.6517 % for QSAR-NN1, 9.1991 % for QSAR-NN2, and 5.8869 % for QSAR-NN3, and for QSAR-MLR models they were 20.1865 % for QSAR-MLR1, 19.3815 % for QSAR-MLR2, and 16.2062 % for QSAR-MLR3).

show abstract

Section: Data Base Collectionmentioning

confidence: 99%

A Comparison of “Neural Networks and Multiple Linear Regressions” Models to Describe the Rejection of Micropollutants by Membranes

2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…Developing a variety of ANN models and then combining them is an appealing technique for increasing this model's robustness. Many academic researchers have looked into combining different neural network models 42,26 to construct BANN models, the training data set was re-sampled using bootstrap re-sampling with replacement 43,44 to create 30 training sets.…”

Section: Bootstrap Aggregated Neural Networkmentioning

confidence: 99%

Prediction of Hourly Global Solar Radiation: Comparison of Neural Networks / Bootstrap Aggregating

Dahmani

Ammi

Hanini

et al. 2023

Kemija U Industriji

View full text Add to dashboard Cite

This research work explores the use of single neural networks and bootstrap aggregated neural networks for predicting hourly global solar radiation. A database of 3606 data points were from the Renewable Energies Development Center, radiometric station 'Shems' of Bouzareah. The single neural networks and bootstrap aggregated neural networks were built together. The precision and durability of neural network models generated with an incomplete quantity of training datasets were improved using bootstrap aggregated neural networks. To produce numerous sets of training data points, the training data was re-sampled utilising bootstrap resampling by replacement. A neural network model was built for each of the data points. The individual neural network models were then combined to produce the bootstrap aggregated neural networks. The experimental and predicted values of global solar radiation were compared, and lower root mean squared errors (68.3968 and 62.4856 Wh m −2 ) were discovered during the testing phases for single neural networks and bootstrap aggregated neural networks, respectively. The results of these models show that the bootstrap aggregated neural networks model is more accurate and robust than single neural networks.

show abstract

“…Nevertheless, only a handful of neural network models exist that can forecast the retention of organic substances in reverse osmosis, forward osmosis, and nanofiltration. (Ammi et al, 2015(Ammi et al, , 2018(Ammi et al, , 2020(Ammi et al, , 2023Ammi, Khaouane, et al, 2021;Ammi, Hanini, et al, 2021;Khaouane et al, 2017;Kratbi et al, 2023;Libotean et al, 2008;Shahmansouri & Bellona, 2013;Yangali-Quintanilla et al, 2009).…”

Section: Introductionmentioning

confidence: 99%

Prediction of the effect of adsorption on the retention of organic compounds by NF/RO using QSPR-ANN

2024

Global NEST Journal

View full text Add to dashboard Cite

Understanding the retention of organic compounds (OCs) is critical for membrane applications in water recycling. The objective of this study was to create an optimized model using Artificial Neural Networks for Quantitative Structure-Property Relationship (QSPR-ANN) to predict the effect of adsorption on the retention of organic compounds (OCs) by nanofiltration (NF) and reverse osmosis (RO). An optimal model (QSPR-ANNoptimal) characterized by a similar structure (13 neurons in the inputs layer, 11 neurons in the hidden layer, and 1 neuron in the output layer) is constructed to predict the effect of adsorption on the retention of organic compounds by membranes. A set of 273 data points was used to test the neural network. the data set was used 70% for training, 15% for validation, and 15% for testing. For the most promising neural network model, the calculated retention values were compared to the experimental retention values, and good correlations were found (the determination coefficient "R2 = 0.9872" and the root mean squared error "RMSE = 2.2743%" for the test phase). This indicates the good robustness of the established QSPR-ANN model and the possibility of predicting the various parameters that characterize the retention of OCs by RO/NF. Sensitivity analysis revealed that the effect of adsorption retention of organic compounds by reverses osmosis and nanofiltration membranes depends more precisely on two important interactions (hydrophobic/adsorption and steric hindrance).

show abstract

A Model Based on Bootstrapped Neural Networks for Modeling the Removal of Organic Compounds by Nanofiltration and Reverse Osmosis Membranes

Cited by 9 publications

References 33 publications

A Comparison of “Neural Networks and Multiple Linear Regressions” Models to Describe the Rejection of Micropollutants by Membranes

A Comparison of “Neural Networks and Multiple Linear Regressions” Models to Describe the Rejection of Micropollutants by Membranes

Prediction of Hourly Global Solar Radiation: Comparison of Neural Networks / Bootstrap Aggregating

Prediction of the effect of adsorption on the retention of organic compounds by NF/RO using QSPR-ANN

Contact Info

Product

Resources

About