PhD Position Hybrid Safe Learning for Inter-Connected Systems

Are you curious how Deep Learning and Online Learning can be effectively combined to create new learning paradigms?

Job description

Online learning algorithms achieve robustness often at the expense of performance, as they are very cautious by design. This, in turn, makes them less practical for problems where speed is of utmost priority. On the other hand, offline learning, such as Deep Learning, often suffers from distribution shifts, lack of training data, and poor adaptability to unseen conditions and new problems. Can we combine these two fundamental learning paradigms to synthesize new learning tools that are both fast and adaptive?

This PhD thesis aims to develop a robust hybrid learning framework that lies at the nexus of online and offline learning. The developed algorithms should be able to benefit from training data, when these are available, and also to learn from real-time, potentially non-IID, streaming data; should be able to track the evolution of key features and achieve model plasticity while avoiding catastrophic forgetting; and should come with interpretable and robust accuracy and performance guarantees.

The designed algorithms will be applied to key problems in the domain of safe learning for interconnected systems, such as 6G and Edge AI platforms and self-driving vehicle vision, in collaboration with industry partners and domain experts.

This PhD thesis is offered in the context of the Marie Curie Doctoral Networks "FINALITY", will be hosted at TU Delft, Department of Computer Science, and will be co-supervised by Prof. George Iosifidis (TU Delft) and Prof. Constantine Dovrolis (University of Cyprus, and Cyprus Institute). The successful candidate will join an academic-industrial consortium working on the foundations and applications of Safe Learning, with secondments at University of Cyprus and other partners.

1. N. Mhaisen, G. Iosifidis, On the Dynamic Regret of Following the Regularized Leader: Optimism with History Pruning, ICML, 2025.
2. G. Iosifidis, N. Mhaisen, D. Leith, Optimistic Learning for Communication Systems, available in Arxiv, 2026.
3. Cameron Ethan Taylor, Shreyas Malakarjun Patil, Constantine Dovrolis: Before Forgetting, There's Learning: Representation Learning Challenges in Online Unsupervised Continual Learning. Trans. Mach. Learn. Res. 2025 (2025)
4. Mustafa Burak Gurbuz, Xingyu Zheng, Constantine Dovrolis: PEAKS: Selecting Key Training Examples Incrementally via Prediction Error Anchored by Kernel Similarity. ICML 2025

Job requirements

• Master’s degree in Computer Science, Machine Learning, Operations Research, Applied Mathematics, or related fields.
• Bachelor degree in Mathematics, Data Science, Computer Science, Electrical Engineering, Operations Research, or related fields.
• Mathematical Foundations: Knowledge of optimization techniques (e.g., LP, CVX, etc), including for/with ML (first order methods, data-driven algorithms, etc).
• Data Foundations: Hands-on experience in data analysis (Python, etc.), experience in evaluation of algorithms, and experience with Deep Learning libraries.
• Excellent command of written and spoken English, subject to TU Delft eligibility criteria.

To thrive as a PhD candidate, it’s crucial to have a strong research mindset driven by curiosity and passion for your topic. Reflecting on your motivation for pursuing a PhD trajectory is essential, as this path involves unique challenges and uncertainties inherent to scientific exploration. Success requires dedication, adaptability, the ability to analyze complex problems, manage your time effectively, innovate and stay resilient under pressure. Combined with the ability and willingness to work independently and collaborate well, these qualities are indispensable for a fulfilling PhD journey. These experiences will build you as an independent researcher, expand your professional network, and pave the way for diverse career paths, inside or outside academia.

Doing a PhD at TU Delft requires English proficiency at a level that ensures the candidate can communicate and interact well, participate in English-taught Doctoral Education courses, and write scientific articles and a final thesis.

TU Delft (Delft University of Technology)

TU Delft is a top international university combining science, engineering and design.

Faculty of Electrical Engineering, Mathematics and Computer Science

The Faculty of Electrical Engineering, Mathematics and Computer Science (EEMCS) brings together three scientific disciplines. Combined, they reinforce each other and are the driving force behind the technology we all use in our daily lives. The faculty offers room for ground-breaking research and provides an innovative environment with excellent labs and facilities.

Conditions of employment

Doctoral candidates will be offered a 4-year period of employment in principle, but in the form of 2 employment contracts. An initial 1.5 year contract with an official go/no go progress assessment within 15 months, followed by an additional contract for the remaining 2.5 years assuming everything goes well and performance requirements are met.

As a PhD candidate you will be enrolled in the TU Delft Graduate School. The TU Delft Graduate School provides an inspiring research environment with an excellent team of supervisors, academic staff and a mentor. The Doctoral Education Programme is aimed at developing your transferable, discipline-related and research skills.

TU Delft offers a customisable compensation package, discounts on health insurance, and a monthly work costs contribution. Flexible work schedules can be arranged.