The candidate holds a master’s degree in one of the following fields (or a closely related discipline): Business Engineering, Applied Economics, Applied Sciences, Engineering, Computer Science, Mathematics, or Informatics.
The ideal candidate combines strong quantitative and problem-solving skills with a genuine interest in developing and applying data-driven methods to address business challenges.
Preferred qualifications include:
1. Knowledge of machine learning (familiarity with large language models is a plus);
2. Programming skills, preferably in Python;
3. Strong communication skills and an excellent command of written and spoken English.
The candidate is expected to apply for a doctoral grant with FWO or Vlaio. Starting date is to be agreed upon, but in principle as soon as possible. You will work under the joint supervision of prof. Wouter Verbeke (KU Leuven, Faculty of Economics and Business, Department of Decision Sciences and Information Management, Research Centre for Information Systems Engineering) and prof. Tim Verdonck (University of Antwerp, Department of Mathematics). Your main office will be at the Research Centre for Information Systems Engineering at KU Leuven (Naamsestraat 69, Leuven, Belgium), but you will be expected to work on a regular basis at the offices of the involved financial institution as well as at the University of Antwerp (1 day per week).
Digital business thrives on secure transaction processes. Yet online transaction processing systems are relentlessly probed for security vulnerabilities arising from system modifications and updates. Despite advanced authentication procedures and network protocols, the Association of Certified Fraud Examiners estimates that a typical organization loses around 5% of its revenues to fraud. Developing powerful fraud detection systems that continuously monitor and learn from data flows is therefore crucial for reducing losses by timely blocking, containing, and preventing malicious behavior.
Much like viruses mutate in response to immunity, fraudsters continuously adapt their tactics and methods in reaction to organizations’ countermeasures. Fraud is thus dynamic, system-dependent, and organization-specific. This poses a critical challenge for applying machine learning: there are typically only a few examples of each specific fraud pattern to learn from. Classic machine learning methods struggle to distinguish these rare cases within the overall population of millions of banking clients.
This project therefore aims to leverage the generalization capabilities of large language models (LLMs) to empower fraud detection systems. By treating transactions as tokens, LLMs can be harnessed to rapidly identify novel fraud patterns and detect or prevent fraud even when only a few known cases are available.
The Research Centre for Information Systems Engineering (LIRIS) coordinates research in the area of information technology and management. This research group is seeking a highly motivated candidate with strong skills and a keen interest in application-oriented academic research in applied machine learning. The selected candidate will conduct a research project on data-driven fraud detection, with the objective of obtaining a PhD degree in Applied Economics. The project will be carried out in close collaboration with a financial institution, ensuring access to real-world data and providing a testbed for evaluating the developed solutions.We offer a full-time position as a doctoral researcher for one year with an extension of up to four years following a positive evaluation. More information on pursuing a Ph.D. at KU Leuven can be found at research.kuleuven.be/en/career/phd, including information on salary and other benefits. More information on the Ph.D. program in Business Economics at the Faculty of Economics and Business can be found at feb.kuleuven.be/research/PhD/businesseconomics/Business.