Calvin Tsay
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Calvin Tsay has a total of 104 co-authors. Here are the top co-authors by number of shared publications:
- Ruth Misener
Department of Computing Imperial College London
7 shared publications - Jan Kronqvist
KTH Royal Institute of Technology
5 shared publications - David Walz
RWTH Aachen University, III. Physikalisches Institut A, Aachen, Germany
3 shared publications - Gabriel D. Patrón
Department of Computing
3 shared publications - Philip Sosnin
Department of Computing Imperial College London United Kingdom
3 shared publications - Antonio Alcántara
Department of Statistics University Carlos III of Madrid Avenida de la Universidad 30 28911 Leganés Spain
2 shared publications
Latest Publications
- 30 September 2026
- Published by American Chemical Society (ACS) in Industrial & Engineering Chemistry Research
- Vol. 65 (40) , 21460-21499
Atoms to Processes: The Role of Artificial Intelligence and Machine Learning in Chemical Engineering
The rapid maturation of artificial intelligence (AI) and machine learning (ML) has catalyzed a profound shift in how chemical engineering problems are formulated, analyzed, and solved. Advances in computing, data availability, and learning algorithms have enabled AI/ML methods to impact applications spanning atomic-scale simulations, materials and catalyst discovery, transport and thermodynamics, separations, process systems engineering, and industrial operations. This Perspective provides an account of recent methodological developments and representative applications, emphasizing how AI/ML tools are being integrated with first-principles models to address challenges of predictive accuracy, data scarcity, extrapolation, interpretability, and model life cycle management. Across domains, a unifying trend is the move away from purely black-box approaches toward hybrid and physics-informed frameworks that explicitly respect conservation laws, thermodynamic consistency, and known structural constraints. These approaches not only improve robustness and reliability but also enable meaningful human–AI collaboration by providing information at an appropriate level of abstraction for the task and decision context. We conclude that AI and ML are not replacing the core principles of chemical engineering; rather, they are amplifying them. As the field advances toward increasingly autonomous, adaptive, and sustainable systems, the thoughtful integration of AI/ML with first-principles understanding and domain expertise will be essential to realizing their full potential across both research and industrial practice.
- 30 September 2026
- Published by American Chemical Society (ACS) in Industrial & Engineering Chemistry Research
- Vol. 65 (40) , 21460-21499
Atoms to Processes: The Role of Artificial Intelligence and Machine Learning in Chemical Engineering
The rapid maturation of artificial intelligence (AI) and machine learning (ML) has catalyzed a profound shift in how chemical engineering problems are formulated, analyzed, and solved. Advances in computing, data availability, and learning algorithms have enabled AI/ML methods to impact applications spanning atomic-scale simulations, materials and catalyst discovery, transport and thermodynamics, separations, process systems engineering, and industrial operations. This Perspective provides an account of recent methodological developments and representative applications, emphasizing how AI/ML tools are being integrated with first-principles models to address challenges of predictive accuracy, data scarcity, extrapolation, interpretability, and model life cycle management. Across domains, a unifying trend is the move away from purely black-box approaches toward hybrid and physics-informed frameworks that explicitly respect conservation laws, thermodynamic consistency, and known structural constraints. These approaches not only improve robustness and reliability but also enable meaningful human–AI collaboration by providing information at an appropriate level of abstraction for the task and decision context. We conclude that AI and ML are not replacing the core principles of chemical engineering; rather, they are amplifying them. As the field advances toward increasingly autonomous, adaptive, and sustainable systems, the thoughtful integration of AI/ML with first-principles understanding and domain expertise will be essential to realizing their full potential across both research and industrial practice.
- Open Access
- 3 September 2026
- Published by Elsevier in Energy Conversion and Management
- Vol. 370
Techno–economic optimisation of hydrogen geological storage in UK depleted gas reservoirs
- 3 September 2026
- Published by Elsevier in Energy Conversion and Management
- Vol. 370
Open AccessTechno–economic optimisation of hydrogen geological storage in UK depleted gas reservoirs
- Open Access
- 17 June 2026
- Published by Elsevier in Computers & Chemical Engineering
- Vol. 213
Graph modeling and optimization in process systems engineering
- 17 June 2026
- Published by Elsevier in Computers & Chemical Engineering
- Vol. 213
Open AccessGraph modeling and optimization in process systems engineering
- 1 June 2026
- Published by Springer Nature
- p. 538-556
Exact Certification of Data-Poisoning Attacks Using Mixed-Integer Programming
- 1 June 2026
- Published by Springer Nature
- p. 538-556
Exact Certification of Data-Poisoning Attacks Using Mixed-Integer Programming
- Top Cited Papers
- 12 March 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in INFORMS Journal on Computing
When Deep Learning Meets Polyhedral Theory: A Survey
In the past decade, deep learning became the prevalent methodology for predictive modeling thanks to the remarkable accuracy of deep neural networks in tasks such as computer vision and natural language processing. Meanwhile, the structure of neural networks converged back to simpler representations based on piecewise constant and piecewise linear functions such as the rectified linear unit (ReLU), which became the most commonly used type of activation function in neural networks. That made certain types of network structure—such as the typical fully connected feedforward neural network—amenable to analysis through polyhedral theory and to the application of methodologies such as linear programming and mixed-integer linear programming for a variety of purposes. In this paper, we survey the main topics emerging from this fast-paced area of work, which brings a fresh perspective to understanding neural networks in more detail as well as to applying linear optimization techniques to train, verify, and reduce the size of such networks. History: Accepted by Andrea Lodi, Area Editor of Design and Analysis of Algorithms—Discrete. Funding: T. Serra was supported by the National Science Foundation (NSF) Division of Information and Intelligent Systems [Grant IIS 2104583], C. Tsay was supported by the Engineering and Physical Sciences Research Council (EPSRC) [Grant EP/T001577/1]. G. Muñoz was supported by the Chilean Research and Development Agency (ANID) [Grant PIA/PUENTE AFB230002].
- 12 March 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in INFORMS Journal on Computing
Top Cited PapersWhen Deep Learning Meets Polyhedral Theory: A Survey
In the past decade, deep learning became the prevalent methodology for predictive modeling thanks to the remarkable accuracy of deep neural networks in tasks such as computer vision and natural language processing. Meanwhile, the structure of neural networks converged back to simpler representations based on piecewise constant and piecewise linear functions such as the rectified linear unit (ReLU), which became the most commonly used type of activation function in neural networks. That made certain types of network structure—such as the typical fully connected feedforward neural network—amenable to analysis through polyhedral theory and to the application of methodologies such as linear programming and mixed-integer linear programming for a variety of purposes. In this paper, we survey the main topics emerging from this fast-paced area of work, which brings a fresh perspective to understanding neural networks in more detail as well as to applying linear optimization techniques to train, verify, and reduce the size of such networks. History: Accepted by Andrea Lodi, Area Editor of Design and Analysis of Algorithms—Discrete. Funding: T. Serra was supported by the National Science Foundation (NSF) Division of Information and Intelligent Systems [Grant IIS 2104583], C. Tsay was supported by the Engineering and Physical Sciences Research Council (EPSRC) [Grant EP/T001577/1]. G. Muñoz was supported by the Chilean Research and Development Agency (ANID) [Grant PIA/PUENTE AFB230002].
- Open Access
Exact Certification of Data-Poisoning Attacks Using Mixed-Integer Programming
This work introduces a verification framework that provides both sound and complete guarantees for data poisoning attacks during neural network training. We formulate adversarial data manipulation, model training, and test-time evaluation in a single mixed-integer quadratic programming (MIQCP) problem. Finding the global optimum of the proposed formulation provably yields worst-case poisoning attacks, while simultaneously bounding the effectiveness of all possible attacks on the given training pipeline. Our framework encodes both the gradient-based training dynamics and model evaluation at test time, enabling the first exact certification of training-time robustness. Experimental evaluation on small models confirms that our approach delivers a complete characterization of robustness against data poisoning.
Open AccessExact Certification of Data-Poisoning Attacks Using Mixed-Integer Programming
This work introduces a verification framework that provides both sound and complete guarantees for data poisoning attacks during neural network training. We formulate adversarial data manipulation, model training, and test-time evaluation in a single mixed-integer quadratic programming (MIQCP) problem. Finding the global optimum of the proposed formulation provably yields worst-case poisoning attacks, while simultaneously bounding the effectiveness of all possible attacks on the given training pipeline. Our framework encodes both the gradient-based training dynamics and model evaluation at test time, enabling the first exact certification of training-time robustness. Experimental evaluation on small models confirms that our approach delivers a complete characterization of robustness against data poisoning.
- Open Access
- 27 November 2025
- Published by Elsevier in Computers & Chemical Engineering
- Vol. 206
A survey and tutorial of reinforcement learning methods in Process Systems Engineering
- 27 November 2025
- Published by Elsevier in Computers & Chemical Engineering
- Vol. 206
Open AccessA survey and tutorial of reinforcement learning methods in Process Systems Engineering
- Open Access
Abstract Gradient Training: A Unified Certification Framework for Data Poisoning, Unlearning, and Differential Privacy
The impact of inference-time data perturbation (e.g., adversarial attacks) has been extensively studied in machine learning, leading to well-established certification techniques for adversarial robustness. In contrast, certifying models against training data perturbations remains a relatively under-explored area. These perturbations can arise in three critical contexts: adversarial data poisoning, where an adversary manipulates training samples to corrupt model performance; machine unlearning, which requires certifying model behavior under the removal of specific training data; and differential privacy, where guarantees must be given with respect to substituting individual data points. This work introduces Abstract Gradient Training (AGT), a unified framework for certifying robustness of a given model and training procedure to training data perturbations, including bounded perturbations, the removal of data points, and the addition of new samples. By bounding the reachable set of parameters, i.e., establishing provable parameter-space bounds, AGT provides a formal approach to analyzing the behavior of models trained via first-order optimization methods.
Open AccessAbstract Gradient Training: A Unified Certification Framework for Data Poisoning, Unlearning, and Differential Privacy
The impact of inference-time data perturbation (e.g., adversarial attacks) has been extensively studied in machine learning, leading to well-established certification techniques for adversarial robustness. In contrast, certifying models against training data perturbations remains a relatively under-explored area. These perturbations can arise in three critical contexts: adversarial data poisoning, where an adversary manipulates training samples to corrupt model performance; machine unlearning, which requires certifying model behavior under the removal of specific training data; and differential privacy, where guarantees must be given with respect to substituting individual data points. This work introduces Abstract Gradient Training (AGT), a unified framework for certifying robustness of a given model and training procedure to training data perturbations, including bounded perturbations, the removal of data points, and the addition of new samples. By bounding the reachable set of parameters, i.e., establishing provable parameter-space bounds, AGT provides a formal approach to analyzing the behavior of models trained via first-order optimization methods.
- Open Access
- 22 October 2025
- Published by American Chemical Society (ACS) in Industrial & Engineering Chemistry Research
- Vol. 64 (44) , 21235-21250
Integrated Design and Scheduling of Hydrogen Processes under Uncertainty: A Quantile Neural Network Approach
The design and scheduling of electrolysis-based hydrogen production plants are subject to uncertainty in future electricity price predictions. Two-stage stochastic programming, which can model this uncertainty, often suffers from high computational costs. In this work, we propose the use of quantile neural networks as surrogates to model the so-called “second stage” for the integrated design and scheduling problem of a hydrogen process. Our surrogate model captures the distribution of the second-stage value function conditioned on electricity price-dependendent first-stage decisions. The neural network surrogate is then embedded into the two-stage stochastic program, enabling solutions to be found without computationally expensive sampling approaches. As the surrogate model outputs a distribution of the second-stage value function, we show this approach further enables joint optimization based on expectation and conditional value at risk, which is a measure of tail risk. Our results show that the inclusion of risk measures leads to higher investments in the electrolyzer and the storage to hedge against high electricity cost scenarios. The surrogate-based approach yields high-quality decisions and requires significantly less computational resources than the conventional sample average approximation.
- 22 October 2025
- Published by American Chemical Society (ACS) in Industrial & Engineering Chemistry Research
- Vol. 64 (44) , 21235-21250
Open AccessIntegrated Design and Scheduling of Hydrogen Processes under Uncertainty: A Quantile Neural Network Approach
The design and scheduling of electrolysis-based hydrogen production plants are subject to uncertainty in future electricity price predictions. Two-stage stochastic programming, which can model this uncertainty, often suffers from high computational costs. In this work, we propose the use of quantile neural networks as surrogates to model the so-called “second stage” for the integrated design and scheduling problem of a hydrogen process. Our surrogate model captures the distribution of the second-stage value function conditioned on electricity price-dependendent first-stage decisions. The neural network surrogate is then embedded into the two-stage stochastic program, enabling solutions to be found without computationally expensive sampling approaches. As the surrogate model outputs a distribution of the second-stage value function, we show this approach further enables joint optimization based on expectation and conditional value at risk, which is a measure of tail risk. Our results show that the inclusion of risk measures leads to higher investments in the electrolyzer and the storage to hedge against high electricity cost scenarios. The surrogate-based approach yields high-quality decisions and requires significantly less computational resources than the conventional sample average approximation.