Augmenting optimization-based molecular design with graph neural networks

Zhang, Shiqiang; Campos, Juan S.; Feldmann, Christian; Sandfort, Frederik; Mathea, Miriam; Misener, Ruth

doi:10.1016/j.compchemeng.2024.108684

Cited by 2 publications

(1 citation statement)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This formulation has been shown to outperform the standard Big-M approach . It is relevant to highlight that OMLT has the capacity to reformulate and optimize both Convolutional Neural Networks (CNN) and Graph Neural Networks (GNN) …”

Section: Overview Of the Toolsmentioning

confidence: 99%

Process Systems Engineering Tools for Optimization of Trained Machine Learning Models: Comparative and Perspective

López-Flores,

Ramírez-Márquez,

Ponce-Ortega

2024

Ind. Eng. Chem. Res.

View full text Add to dashboard Cite

This article studies the relevance of innovative Process Systems Engineering (PSE) tools that can reformulate trained machine learning models that are driven by advances in computational technologies, showcasing a pivotal transformation in chemical engineering methodologies. The article also delves into how trained machine learning models are reformulated and optimized to refine engineering decisions as it provides a novel analysis of tools to develop machine learning models by reformulating them, and optimizing them in PSE, thus highlighting their significance and applications. It offers a comprehensive comparison of several cutting-edge tools, including JANOS, MeLOn, ENTMOOT, reluMIP, OptiCL, Gurobi Machine Learning, OMLT, and PySCIPOpt-ML, highlighting their distinct abilities for performance and decision-making. Furthermore, challenges related to the explicit formulation of the main machine learning models are discussed. Guidance is provided to select the appropriate tool according to users' requirements. Additionally, a comparative study of the tools is presented using a case study to analyze and compare the size and type of formulations, the optimal solution, and the computation times.

show abstract

Section: Overview Of the Toolsmentioning

confidence: 99%

Process Systems Engineering Tools for Optimization of Trained Machine Learning Models: Comparative and Perspective

López-Flores,

Ramírez-Márquez,

Ponce-Ortega

2024

Ind. Eng. Chem. Res.

View full text Add to dashboard Cite

show abstract

Glass Transition Temperature Prediction of Polymers via Graph Reinforcement Learning

Dong,

Li,

Liu

2024

Langmuir

View full text Add to dashboard Cite

An expansive array of graph-based models has been utilized for accurate prediction of the structure−property relation of polymers. However, these approaches notably underutilize unsupervised structural information. Concentrating on the domain of heterocyclic polymers, particularly polyimides, this study delves into the glass transition temperature (T g ) prediction, aiming to fully exploit the potential within both the global and local structures of molecules. To achieve this, a graph reinforcement learning framework termed Molecular Structural Regularized Graph Convolutional Network with Reinforcement Learning (MSRGCN-RL) is proposed. Experimental results highlight the crucial role of both global and local structural regularization in precise T g prediction. Concurrently, optimization of MSRGCN training through RL proves essential. This research leads the way in integrating Graph Neural Networks (GNNs) with reinforcement learning methodologies for the property prediction of polymers.

show abstract

Augmenting optimization-based molecular design with graph neural networks

Cited by 2 publications

References 44 publications

Process Systems Engineering Tools for Optimization of Trained Machine Learning Models: Comparative and Perspective

Process Systems Engineering Tools for Optimization of Trained Machine Learning Models: Comparative and Perspective

Glass Transition Temperature Prediction of Polymers via Graph Reinforcement Learning

Contact Info

Product

Resources

About