no code implementations • 21 May 2020 • Artur M. Schweidtmann, Dominik Bongartz, Daniel Grothe, Tim Kerkenhoff, Xiaopeng Lin, Jaromil Najman, Alexander Mitsos
Often, Gaussian processes are trained on datasets and are subsequently embedded as surrogate models in optimization problems.
no code implementations • 1 Jun 2022 • Jan G. Rittig, Martin Ritzert, Artur M. Schweidtmann, Stefanie Winkler, Jana M. Weber, Philipp Morsch, K. Alexander Heufer, Martin Grohe, Alexander Mitsos, Manuel Dahmen
We propose a modular graph-ML CAMD framework that integrates generative graph-ML models with graph neural networks and optimization, enabling the design of molecules with desired ignition properties in a continuous molecular space.
no code implementations • 23 Jun 2022 • Jan G. Rittig, Karim Ben Hicham, Artur M. Schweidtmann, Manuel Dahmen, Alexander Mitsos
We train the GNN on a database including more than 40, 000 AC values and compare it to a state-of-the-art MCM.
no code implementations • 25 Jul 2022 • Laura Stops, Roel Leenhouts, Qinghe Gao, Artur M. Schweidtmann
In particular, the graph neural networks are implemented within the agent architecture to process the states and make decisions.
no code implementations • 25 Jul 2022 • Gabriel Vogel, Lukas Schulze Balhorn, Edwin Hirtreiter, Artur M. Schweidtmann
SFILES is a text-based notation for chemical process flowsheets.
no code implementations • 25 Jul 2022 • Jan G. Rittig, Qinghe Gao, Manuel Dahmen, Alexander Mitsos, Artur M. Schweidtmann
Molecular property prediction is of crucial importance in many disciplines such as drug discovery, molecular biology, or material and process design.
no code implementations • 27 Jul 2022 • Artur M. Schweidtmann, Jan G. Rittig, Jana M. Weber, Martin Grohe, Manuel Dahmen, Kai Leonhard, Alexander Mitsos
We recommend using sum pooling for the prediction of properties that depend on molecular size and compare pooling functions for properties that are molecular size-independent.
no code implementations • 1 Aug 2022 • Gabriel Vogel, Lukas Schulze Balhorn, Artur M. Schweidtmann
We pre-train our model on synthetically generated flowsheets to learn the flowsheet language grammar.
no code implementations • 26 Oct 2022 • Edwin Hirtreiter, Lukas Schulze Balhorn, Artur M. Schweidtmann
Thereafter, the model is fine-tuned leveraging transfer learning on real P&IDs.
no code implementations • 7 Feb 2023 • Qinghe Gao, HaoYu Yang, Shachi M. Shanbhag, Artur M. Schweidtmann
We thus propose to utilize transfer learning for process design with RL in combination with rigorous simulation methods.
no code implementations • 7 Feb 2023 • Lukas Schulze Balhorn, Edwin Hirtreiter, Lynn Luderer, Artur M. Schweidtmann
Artificial intelligence has great potential for accelerating the design and engineering of chemical processes.
no code implementations • 15 Aug 2023 • Qinghe Gao, Artur M. Schweidtmann
The transformation towards renewable energy and feedstock supply in the chemical industry requires new conceptual process design approaches.
no code implementations • 2 Dec 2023 • Tom McDonald, Calvin Tsay, Artur M. Schweidtmann, Neil Yorke-Smith
ReLU neural networks have been modelled as constraints in mixed integer linear programming (MILP), enabling surrogate-based optimisation in various domains and efficient solution of machine learning certification problems.
no code implementations • 5 Dec 2023 • Lukas Schulze Balhorn, Marc Caballero, Artur M. Schweidtmann
The model achieves a top-1 accuracy of 80% and a top-5 accuracy of 84% on an independent test dataset of synthetically generated flowsheets.
no code implementations • 29 Jan 2024 • Michal Tkáč, Jakub Sieber, Lara Kuhlmann, Matthias Brueggenolte, Alexandru Rinciog, Michael Henke, Artur M. Schweidtmann, Qinghe Gao, Maximilian F. Theisen, Radwa El Shawi
Machine Learning (ML) techniques are encountered nowadays across disciplines, from social sciences, through natural sciences to engineering.
no code implementations • 4 Mar 2024 • Yidong Zhao, Joao Tourais, Iain Pierce, Christian Nitsche, Thomas A. Treibel, Sebastian Weingärtner, Artur M. Schweidtmann, Qian Tao
Deep learning (DL)-based methods have achieved state-of-the-art performance for a wide range of medical image segmentation tasks.