PhD on automated feature discovery and data-driven modelling for chemical processes KU Leuven in Belgium

PhD opportunity at KU Leuven in automated feature discovery and data-driven modelling for chemical processes.

The research group at KU Leuven works at the intersection of chemical engineering, process systems engineering, and artificial intelligence. The project is part of EL4CHEM – Efficient Learning for Chemical Applications and focuses on developing automated methods to identify the most informative features and build reliable data-driven models for complex chemical and pharmaceutical processes.

The PhD researcher will study how to select and generate the right variables for modelling when process datasets contain many possible inputs, such as operating conditions, sensor measurements, material properties, molecular descriptors, and engineered process variables. Because experimental data are often scarce, noisy, and expensive, the project aims to make feature discovery and model identification more intelligent, robust, and interpretable.

Research may include automated feature generation and feature selection, sparse and interpretable machine-learning models, nonlinear feature interactions, dimensionality reduction, automated comparison of model structures, incorporation of physical and chemical knowledge into selection workflows, uncertainty and robustness analysis, explainable AI, and development of automated modelling workflows for different industrial use cases.

The methods will be tested on real applications from the chemical, pharmaceutical, and manufacturing sectors through the EL4CHEM consortium. The project is therefore well suited to a candidate who enjoys combining methodological AI research with practical industrial modelling challenges.

Candidate profile: the advert seeks applicants with a background in chemical engineering, process engineering, applied mathematics, data science, computer science, or a related field. Python skills and prior experience with machine learning, statistical modelling, feature selection, optimisation, or process modelling are advantageous. A strong interest in applying machine learning to chemical-engineering problems is essential, together with motivation, independence, and enthusiasm for interdisciplinary research.

Offer: a full-time research position for one year, with the possibility of extension up to four years depending on performance and available funding. The position offers an international and multidisciplinary environment, opportunities for publication, collaboration with industrial partners, and development of new data-driven modelling methodologies with direct industrial relevance.

Application and contact: interested candidates are invited to apply via the KU Leuven job link. For more information, contact Prof. dr. Mumin Enis Leblebici at [email protected].

Location: Leuven, Belgium.

Deadline: 2026-09-30.