Job description
PhD Thesis on Algorithmic Fairness in Data-Driven Decision Making Project Description Position Description Are you interested in ethical AI and eager to develop solutions to the challenges posed by biased decision making in machine learning (ML)? This PhD position offers a unique opportunity to contribute to cutting-edge research in algorithmic fairness, ensuring that automated decision-making systems produce fair and equitable outcomes. Project Description Machine learning algorithms are increasingly influencing decisions that matter in society – affecting energy pricing, college admissions, loan approvals, and more. While these systems can increase efficiency, they often inherit and amplify social biases that can raise serious ethical, legal, and social issues. In practice, automated decision-making can have dynamic feedback effects on the system itself that persist over time. For example, ML-driven energy pricing models adjust costs based on demand, supply, and user behavior. However, these mechanisms may inadvertently disadvantage vulnerable groups by charging higher costs during peak hours, thus reinforcing existing gaps in energy access. To counteract this effect, tools such as optimal control, distributed control and optimization can be used to enforce equity constraints and ensure that efficiency and cost-effectiveness are balanced with equity. Despite recent efforts to reduce bias, there is still a fundamental gap in understanding the long-term feedback dynamics of biased decision-making systems. Your Role As a PhD candidate, you will develop innovative techniques and tools to analyze and mitigate bias propagation in automated systems. Your research focuses on closed-loop interactions between decision-making algorithms and user behavior across a variety of applications, including energy allocation recommender systems e-commerce online advertising By applying concepts from optimization, optimal control theory, nonlinear control, and networked systems, you will study how biases evolve over time and explore strategies for balancing fairness and efficiency in complex real-world environments. Tasks develop strategies to minimize bias propagation and promote fairness in different application domains such as energy allocation, recommender systems, e-commerce, and online advertising. Research the literature on algorithmic fairness in automated decision-making systems; analyze key factors contributing to bias propagation in ML algorithms; analyze computational efficiency and efficiency-fairness trade-offs of recommendation strategies; empirically validate designed algorithms on real datasets; disseminate your research results in international and peer-reviewed journals and conferences; successfully write and defend a dissertation based on your research results; take on masters’ student supervision and internships and other educational tasks
offer requirements
We are looking for a candidate who meets the following requirements: you are an enthusiastic, open-minded young researcher who wants to make a difference in your research field; you have experience or background in systems and control, mathematics, physics, machine learning, data-driven modeling, signal processing; you preferably already have a Master’s degree in Systems and Control, Mechanical or Electrical Engineering, (Applied) Physics or (Applied) Mathematics; you are working in a team with a team of researchers; you are a member of a team of researchers who are working on a research project. Master’s degree in mathematics; works well in a team and is interested in socio-technical applications; good programming skills and experience; good communication skills and cooperative attitude in a research team; intellectually honest; innovative and ambitious, hard working and persistent; good command of English (Dutch not required).
offer benefits
A rewarding career in a dynamic and ambitious university, in an interdisciplinary environment and within an international network. You will work on a beautiful green campus within walking distance of the central train station. In addition, we offer you a full-time job for four years, with a mid-term evaluation (go/no-go) after nine months. Teaching experience – to develop your teaching skills, you will spend 10% of your working time on teaching assignments. Interdisciplinary research – Work at the intersection of machine learning, control theory and AI ethics at a dynamic and ambitious university. Postgraduate courses and training – Take specialized courses at the Dutch School of Systems and Control. Collaboration Opportunities – Collaborate with industry partners in the BrainPort region and work with leading researchers from around the world. Dynamic Research Environment – Join the dynamic Control Systems group within the Department of Electrical Engineering and work with PhD students, postdocs, and faculty (~40 researchers) on a variety of real-world control applications. Compensation and benefits (e.g., pension plan, paid pregnancy and maternity leave, partially paid parental leave) in line with the Dutch Universities Collective Labor Agreement at level P (min. €2,901, max. €3,707). 8.3% year-end bonus and 8% annual leave pay. High-quality training programs and other support to help you develop into a self-aware, autonomous scientific researcher. At TU/e, we encourage you to learn on your own. State-of-the-art technical infrastructure, on-campus nursery and sports facilities. Subsidized commuting, work-from-home and internet costs. Employee immigration panel and tax compensation program for international candidates (30% facilitation).
To apply for this job please visit tue.nl.