Adem R.N. Aouichaoui

Adem R.N. Aouichaoui

ML for Molecular properties

Publications

You can also find my articles on my Google Scholar profile.

Towards Self-Consistent Graph Neural Networks for Predicting the Ideal Gas Heat Capacity, Enthalpy, and Entropy

Published in Computer Aided Chemical Engineering, 2024

In this paper we present one hybrid modelling configuration that combines graph neural networks with a fundamental equation for calculation the heat capacity of organic compounds.

Recommended citation: Aouichaoui, A. R., Müller, S., & Abildskov, J. (2024). Towards Self-Consistent Graph Neural Networks for Predicting the Ideal Gas Heat Capacity, Enthalpy, and Entropy. In Computer Aided Chemical Engineering (Vol. 53, pp. 2833-2838). Elsevier. https://doi.org/10.1016/B978-0-443-28824-1.50473-7

Scoping and Identifying Data-Driven Optimization Prospects in the Danish Processing Industry

Published in Computer Aided Chemical Engineering, 2024

In this paper we present a new framework for developping ditital twins intended for process monitoring and optimization.

Recommended citation: Aouichaoui, A. R., Ernstsson, B. B., Jul-Rasmussen, P., Iversen, N. H., Vermue, L., & Huusom, J. K. (2024). Scoping and Identifying Data-Driven Optimization Prospects in the Danish Processing Industry. In Computer Aided Chemical Engineering (Vol. 53, pp. 451-456). Elsevier. https://doi.org/10.1016/B978-0-443-28824-1.50076-4

Energy- and process real-time optimization through hybrid modeling - a case from Viking Malt A/S

Published in Computer Aided Chemical Engineering, 2024

In this paper we apply a hybrid modelling to monitor the germination process in the malting process.

Recommended citation: Jul-Rasmussen, P., Aouichaoui, A. R., Korzepa, M., Engelbrecht, K., Kongsgaard, K., & Huusom, J. K. (2024). Energy-and process real-time optimization through hybrid modeling-a case from Viking Malt A/S. In Computer Aided Chemical Engineering (Vol. 53, pp. 1645-1650). Elsevier. https://doi.org/10.1016/B978-0-443-28824-1.50275-1

S-GNN: State-Dependent Graph Neural Networks for Functional Molecular Properties

Published in Computer Aided Chemical Engineering, 2023

In this paper we extended the applicability of graph neural networks to temperature dependent properties. This extension was tested on 13 different properties.

Recommended citation: Aouichaoui, A. R., Cogliati, A., Abildskov, J., & Sin, G. (2023). S-GNN: State-Dependent Graph Neural Networks for Functional Molecular Properties. In Computer Aided Chemical Engineering (Vol. 52, pp. 575-581). Elsevier. https://doi.org/10.1016/B978-0-443-15274-0.50091-3

Application of interpretable group-embedded graph neural networks for pure compound properties

Published in Computers & Chemical Engineering, 2023

In this paper we apply two new interpretable graph neural networks to 30 different pure componenent properties. This is so far the largest study of GNN on pure componenent poperties.

Recommended citation: Aouichaoui, A. R., Fan, F., Abildskov, J., & Sin, G. (2023). Application of interpretable group-embedded graph neural networks for pure compound properties. Computers & Chemical Engineering, 176, 108291. https://doi.org/10.1016/j.compchemeng.2023.108291

Application of Outlier Treatment Towards Improved Property Prediction Models

Published in Computer Aided Chemical Engineering, 2022

In this paper we apply outlier treatment to improve a set of 18 group-contribution based property models.

Recommended citation: Aouichaoui, A. R., Mansouri, S. S., Abildskov, J., & Sin, G. (2022). Application of Outlier Treatment Towards Improved Property Prediction Models. In Computer Aided Chemical Engineering (Vol. 51, pp. 1357-1362). Elsevier. https://doi.org/10.1016/B978-0-323-95879-0.50227-7

On The Interpretability of Graph Neural Networks in QSPR Modeling

Published in Computer Aided Chemical Engineering, 2022

In this paper we explore a new geaph neural network that uses groups as nodes and combine it with the attention mechanism to highlight important substructures.

Recommended citation: Fan, F., Aouichaoui, A. R., & Sin, G. (2022). On The Interpretability of Graph Neural Networks in QSPR Modeling. In Computer Aided Chemical Engineering (Vol. 51, pp. 1393-1398). Elsevier. https://doi.org/10.1016/B978-0-323-95879-0.50233-2

Molecular representations in deep-learning models for chemical property prediction

Published in Computer Aided Chemical Engineering, 2022

In this paper we compare three different techniques for modelling the enthalpy of formation: group-contribution, QSPR and graph neural networks.

Recommended citation: Aouichaoui, A. R., Fan, F., Mansouri, S. S., & Sin, J. A. G. (2022). Molecular representations in deep-learning models for chemical property prediction. In Computer Aided Chemical Engineering (Vol. 49, pp. 1591-1596). Elsevier. https://doi.org/10.1016/B978-0-323-85159-6.50265-7

DeepGSA: Plant Data-Driven Global Sensitivity Analysis using Deep Learning

Published in Computer Aided Chemical Engineering, 2022

In this paper, we developped a Matlab tool for global sensitivity analysis for deep neural netwrks (DNN).

Recommended citation: Aouichaoui, A. R., Al, R., & Sin, G. (2022). DeepGSA: Plant Data-Driven Global Sensitivity Analysis using Deep Learning. In Computer Aided Chemical Engineering (Vol. 49, pp. 1759-1764). Elsevier. https://doi.org/10.1016/B978-0-323-85159-6.50293-1

Combining Group-Contribution concept and graph neural networks toward interpretable molecular property models

Published in Journal of Chemical Information and Modeling, 2022

In this paper we develop two new interpretable graph neural networks by integrating groups knwn from group-contribution models as nodes and adding the attention mechanism to highlight important groups.

Recommended citation: Aouichaoui, A. R., Fan, F., Mansouri, S. S., Abildskov, J., & Sin, G. (2023). Combining Group-Contribution concept and graph neural networks toward interpretable molecular property models. Journal of Chemical Information and Modeling, 63(3), 725-744. https://doi.org/10.1021/acs.jcim.2c01091

Uncertainty estimation in deep learning‐based property models: Graph neural networks applied to the critical properties

Published in AIChE journal, 2022

In this paper we investigate three model agnostic approaches to uncertainty estimation of graph neural networks: ensemble, bootstrap and monte carlo dropout. these techniques were benchmarked on two GNN models and three pure component properties: the critical temperature, pressure and volume.

Recommended citation: Aouichaoui, A. R., Mansouri, S. S., Abildskov, J., & Sin, G. (2022). Uncertainty estimation in deep learning‐based property models: Graph neural networks applied to the critical properties. AIChE Journal, 68(6), e17696. https://doi.org/10.1002/aic.17696

Comparison of group-contribution and machine learning-based property prediction models with uncertainty quantification

Published in Computer Aided Chemical Engineering, 2021

In this paper, we benchmark and compare three models and techniques for uncertainty quantification applied to a group-contribution model for the lower flammability limit.

Recommended citation: Aouichaoui, A. R., Al, R., Abildskov, J., & Sin, G. (2021). Comparison of group-contribution and machine learning-based property prediction models with uncertainty quantification. In Computer Aided Chemical Engineering (Vol. 50, pp. 755-760). Elsevier. https://doi.org/10.1016/B978-0-323-88506-5.50118-2

Analysis and evaluation of periodic separations using COPS trays

Published in Chemical Engineering Transactions, 2018

In this paper we evaluate the performance of cyclically operated perforated sheet (COPS) of periodic stripping of a dilute aqueous ammonia solutions with air.

Recommended citation: Nielsen A., Alvarez E.C., Carlsen N., Azizi H., Jorgensen S.B., Abildskov J., 2018, Analysis and evaluation of periodic separations using cops trays, Chemical Engineering Transactions, 69, 733-738 https://doi.org/10.3303/CET1869123s